India and The Computer A study of Planned Development 0195627350

Citation preview

INDIA AND THE COMPUTER A Study of Planned Development

C. R. Subramanian

DELHI

OXFORD UNIVERSITY PRESS BOMBAY CALCUTTA MADRAS

1992

•S ¥- Oxford University Press, Walton Street, Oxford OX2 6DP New York Toronto Delhi Bombay Calcutta Madras Karachi Petaling Jaya Singapore Hong Kong Tokyo Nairobi Dar es Salaam Melbourne Auckland and associates in Berlin Ibadan

©Oxford University Press 1992 SBN 0 19 562735 0

Printed at Rekha Printers Pvt Ltd, New Delhi 110020 and published by S. K. Mookerjee, Oxford University Press YMCA Library Building, Jai Singh Road, New Delhi 110001

To my wife Parvathi for her unfailing support

University Libraries C a r n e g i e Lleilon University Pittsburgh PA 1 5 2 1 3 - 3 8 9 0

'

—

Contents

List of Tables Preface Acknowledgements A Note on Units of Currency and Value

J* X1U XVI X1X

I

Sclf-Reliant Development Fails (1968-78)

II

Quest for New Policies (1978-84)

38

1

III

The Policy of Rajiv Gandhi (1984-90)

60

IV

Peripherals Pose Difficulties

V

Mainframes - A Late Awakening

109

VI

Software - Dreams of Riches

133

VII VIII

The Exit of IBM Indigenous Development — The ECIL Experience

161

IX

CMC Ltd. - A Qualified Success

208

X

Technology and Applications Development

222

XI

The Private Sector Takes the Lead

274

XII

A Look to the Future

280

83

Appendices A.

Brazil and China

^15

B.

Policy Guidelines for the Import of Computers - 1975

33^

C.

The Minicomputer Policy - 1978

336

D.

November 1984 Policy

339

viii

Contents

E.

Computer Imports for Software Exports

F.

1974 Policy Software Export Policy - 1981

G.

Software Policy - 1986

List of Abbreviations Bibliography Index of Names Subject Index

List of Tables

1.1

Bhabha Committee Projections

1.2

Computer Applications (1968) Narasimhan Tabulation

1.3

Electronic Markets, USA (1966 forecasts)

1.4

US Federal Funds for Research and Development (1966)

1.5

R&D Expenditure in Industry

1.6

Minicomputer Panel - Estimate of requirements - (Fifth Five Year Plan)

1.7

Computers - DOE Expenditure A Summary up to 1978

1.8

Production (1976-9)

1.9

Computer Installations (May 1978)

1. 10

Distribution of ECIL Computers

1.11

Import of Computer Systems (1976-81)

2.1

Minicomputer Production (1980-2)

2.2

Computer Demand Profile (1982-90)

2.3

Estimated Demand for Peripherals

2.4

Seventh Plan Projection-Computer Demand

2.5

Seventh Plan Projection-Computer Demand (1985-90)

4 f. o

10

11 11 19 30 31 31 32 32 42 45 45 46 47

List of Tables

X

2.6

Seventh Plan Projection-Demand Projections for Computers

2.7

47

Import of Computers - Government Account (1980-4)

52

2.8

Price Analysis (1983)

55

2.9

Import and Production (1980-4)

56

3.1

Production of Computers and Peripherals

75

4.1

Peripheral Manufacturing Licences

85

4.2

Peripheral Licences Granted to BEL

87

4.3

Sondhi Committee Report - List of Licences up to 1979

93

4.4

Peripheral Industry (1983)

94

4.5

Demand Projection for Computer Peripherals

4.6

Summary of Demand Projections Peripherals

4.7

4.9 4. 10

6.2 6.3

99

Peripherals - Licensed Capacity/Foreign Collaborations

101

Production of Peripherals (1985-90)

103

Major Indian Peripheral Manufacturers (1991)

6. 1

98

Peripherals - Licensing Recommendations Rao Committee

4.8

97

104

Conditions for the Import of Computers for Software Export

136

Software Exports (1975-8)

139

Computer Imports for Software Exports (1981-4)

141

6.4

Software Production

142

6.5

Ten Independent US Software CompaniesSales - R&D Expenses - (1987)

152

6.6

Software Domestic and Export

153

6.7

Indian Software Industry - (1990-1)

154

7. 1

IBM 1401 Reconditioning Programme

162

List of Tables 7.2

IBM - Indian Operations - Financial

163

7.3

Summary IBM Operations - Foreign Exchange

7.4

Outflow IBM - Clients Served-Persons Employed

8.1

ECIL's Projections

8.2

The Proposed Range of EC1L Computers

8.3

ECIL Production up to 1978

8.4

Total ECIL Computer Sales to Private Parties

193

8.5

ECIL Computer Deliveries up to (1986-7)

194

8.6

ECIL - Invcstmcnts/Profit/Loss (1970-87)

196

8.7

System Software Developed by ECIL

198

8.8

ECIL Computer Division Performance - Sales

204

9.1

CMC Ltd. -Turnover (1989-90)

9.2

CMC - Value Added/Value Applied

216

9.3

CMC - Incomc/Expcnse/Profit

215

9.4

CMC Ltd. - Financial Position (31 March 1990)

217

9.5

Percentage Profit before Tax (1987-90)

218

9.6

Systems and Equipment under CMC Support (1990)

219

9.7

CMC Seventh Plan Targets Revenue Projections (1989-90)

219

10.1

Technology DevelopmentCompleted Projects (1971-84)

225

10.2

Technology Development - Ongoing

10.3

Projects (1984) NCSDCT 'Paper' Output Count

10.4

NIC Financial Outlays

10.5

Manpower DevelopmentComputer Requirements (1980-5)

10.6

Computer Manpower Requirements (1985-90)

165 171 178 180 185

214

226 234 239 250 251

List of Tables

xii 10.7

Eighth Plan Manpower Requirements and Availability

10.8

253

Industiial Electronics Promotion Programmes

256

10.9

Grant-In-Aid for Computer-Aided Design

259

10. 10

Computer-Aided Management Outlays

261

10. 11

Knowledge-based Computer System Outlays

262

10. 12

Technology and Applications Development Summary of Expenditures (1985-90)

10. 13

Technology and Applications Development Summary of the Use of Funds (1985-90)

10. 14 11. 1 11.2

267 268

UNDP Assistance for Computer Technology Development

271

Top Twenty Computer Companies in India (1991)

276

Information Technology Activity in India - A Rough Outline (1991)

277

12. 1

US Computer Industry-Scctor-wise Salcs/Profits/R&D

287

12.2

Typical Salcs/Profits/R&D Outlays US Computer Industry

288

12.3

Licences Issued

306

12.4

Summary of Industrial Licences, Letters of Intent, and Registrations up to 1987 end

A 1 A 2

308

Brazilian Computer Industry Performance (1981-6)

320

Brazilian Computer Products - Import/Export

321

Preface

Nehru's will prevailing, independent India chose the path of planned, regulated, controlled development. The public sector was to dominate the commanding heights of the economy; the private sector allowed a role in a tightly regulated environment. Nehru also wanted the nation to acquire a 'scientific temper', so he encouraged science and scientists. Mrs Gandhi continued this chosen path, though several events of a geo-political character impeded progress, as did the political leaders ot a lesser mettle who fell prey to the tcmpations of office and who slowly subverted the bureaucracy and the system to their personal advantage. Until the 1960s, the area of electronics including computers, was the preserve of the Directorate General of Technical Development (DG1 D) under the Ministry of Industry. They had the power to promote, to control, and to regulate the industry. The Department of Defence Supplies was later vested with authority to oversee this industry till the formation ot the Electronics Commission and the Department of Electronics in 1970. Scientists sought control over this glamour industry, and events like the war with China, and the interest of Uomi Bhabha (the eminent nuclear scientist) in electronics, brought them the power they sought. The scientists believed in self-reliant development as the path to progress. As the Electronics Commission and the Department ot Electronics was staffed mainly by the scientists and engineers drawn from the Atomic Energy Establishment and the Tata Institute ot Fundamental Research, this belief in self-reliant development was natural. They had already demonstrated their ability in the nuclear field. But a technology and an industry as diverse as electronics, which was by then already

xiv

Preface

making phenomenal progress could not wait for Indian scientists to catch upMany of the top scientists had thought that if they could enter the government they could influence policies, but the politician is not easily converted. The talk of 'scientific temper' is fine from a platform, but staying in power needs the support of trade and industry. Elections have to be won and only business interests and traders can provide the support — not the scientists, however eminent they may be. Policies with ideological underpinnings are only for the book — they are not practical in the politics of government in the Indian milieu. The domination of scientists over policy formulation gave way in the 1980s to the lobby pressures from trade and business groups. The anxiety to show India off as a technologically advanced nation, even if be with bought hardware and software, and also to derive the advantages of computerization in certain spheres, would not brook the delays and difficulties inherent in a self-reliant development policy. Rajiv Gandhi was to allow the import of computers to promote computerization. He sought to deregulate and decontrol the industry and reduce the predominance of the public sector. But he was voted out of power, and later, tragically assassinated. The importance of information technology and microelectronics for national development need hardly be stressed. A developing country needs resources for designing, building and exploiting information tech­ nology products. The developed world and the multinational companies (MNCs) who arc part of that world, set their own terms to part with technology. Likewise, the developing countries have their own agenda. There arc many schools of thought on the subject of relations between developing countries and the developed world and the MNCs as regards the subject of technology transfer. The so-called 'dependency' school believes that developing countries tied to international capitalism will remain economically dependent on the developed nations for all time to come, in spite of internal skills and significant economic development. Developing countries also fear that if the unchecked flow of technology is permitted, cither by direct foreign investment or by passive reliance on continuous licensing, the ability to self-generate know-how will be lost. Another school of thought believes that developing countries can strike bargains with the developed world and the MNCs. Some nations, with the will to do so, will persevere. Some already have. This book presents India's struggles in developing policies and in implementing them in the field of computers. It is not a book that theorises

xv

Preface

on policies or attempts to apply various principles of development economics or philosophies to the Indian computer scene. It attempts to present the reader with a historical review of policies, and their laggard implementation as they evolved. The book may also serve as a case study of the process of planned development of a fast-moving technology in a developing country. The planners in India spent a lot of time, effort, and money in forecasting demand for each of the five year national plans. These plans and forecasts have all gone awry as coherent policies could never be framed. The import policies, tariffs, fiscal measures, licensing procedures, etc. were never in consonance with the broad aims of the policies laid down from time to time. And there never was a computeriza­ tion policy as such — apparently, it was not politically expedient to have one. The socialist world has now been forced to move towards an openmarket economy. The Indian experiment of a'mixed economy based on licensing, controls and regulation has not proved successful. It is hoped that this study of the planned development of computer technology in India would serve to highlight the weaknesses of such a system. C. R. SUBRAMANIAN BANGALORE

Acknowledgements

I would like to record my gratitude to Dr C.N.R. Rao, and the Indian National Science Academy for their support. In 1985,1 approached them for financial assistance to write a book relating to policies in the field of Electronics in India. Dr Rao felt that a historical review of policies in the field of technology development would be useful. The Indian National Science Academy came forward with a generous grant which was dis­ bursed through the Department of Continuing Education, Indian Institute of Science, Bangalore. The Department of Continuing Education and the Accounts Officer of the Indian Institute Science ensured the prompt disbursement of the funds. Mr S.R. Vijayakar, who served in 1985 as the Secretary, Department of Electronics, arranged for a grant from the Department of Electronics. The Management of Bharat Electronics Ltd., Bangalore, an institution which I had the privilege of serving for over twenty-four years, provided a generous grant and other assistance when­ ever needed. I must thank Mr Aditya Birla of the Birlas for his generous support, and the Senior Executive President of Grasim Industries Ltd., Mr M.C. Bagrodia, for his help, and for the financial support from Philips, India (Peico Electronics and Elcctricals Ltd.) and Samtcl (India) Ltd. The Secretaries of the Department of Electronics, and their staff spared valuable time to discuss various aspects of policies in the field of com­ puters. The Department also provided data for this work. Though they were aware that a critical review may follow, they extended full co-opera­ tion. A book of this nature would not have been possible without their support and it gives me great pleasure to record my gratitude. It was a pleasure to discuss various policy aspects in the field of

A cknowledgements

xvn

computers with Prof.R. Narasimhan, Senior Professor, Tata Institute of Fundamental Research, who made available some of his reports on the subject, and who has kindly read the book and made useful suggestions for its improvement. I must also thank Mr Tejcshwar Singh for his suggestions. His guidance has been of great help. 1 derived valuable insight into the philosophies prevailing in the early 1970s from Dr A.S. Rao. Age has not diminished Dr Rao's enthusiasm or his commitment for self-reliant development. Though one may not share some of his viewpoints, it must be said that India needs many more such dedicated individuals. Mr B.S. Prabhakar, Chairman and Managing Director of ECIL, made available statistical and other data concerning ECIL. Dr P.P. Gupta, Chairman and Managing Director of CMC Ltd., made available all the published annual reports of the CMC Ltd. I must thank both these public sector companies for their co-operation and assistance. I have received assistance and help from many others and it would be difficult to name all of them but I must specifically thank Dr S. Srikantan, who till recently served as the Managing Director of the Karnataka State Electronics Development Corporation Ltd. and Prof.S. Sampath, a former Member of the Union Public Service Commission and an eminent educationist. Ms Indira Soman, Library Director, the United States Infor­ mation Service, Madras, was kind enough to retrieve information tor me on various subjects. I must thank the Hon. Speaker of the Lok Sabha for permission to reproduce extracts from various parliamentary publications and debates. The quotations and extracts have been individually noted/cited in acknow­ ledgement. The journal (now entitled Electronics Information & Planning) pub­ lished by the Information Planning and Analysis Group of the Department of Electronics, has been an important and authoritative source of infor­ mation and I would like to acknowledge my debt to this journal. I am grateful to the editors of Dataquest and Computers Today tor their permission to use several extracts from these magazines. I would like to thank St Martin's Press, New York and Pinter Publishers, London for their permission to use extracts from Paulo Bastos Tigre's

Technology and Competition in the Brazilian Computer Industry, Sage Publications India Private Limited for the quotations from Sharad S. Marathc's Regulation and Development—India's Policy Experience of Controls over Industry, Macmillan for use of extracts from Richard Thomas DcLamcrtcr's Big Blue—IBM's Use and Abuse of Power, the

xviii

A cknowledgements

Research Policy Institute, University of Lund, Sweden for use of Per 0yvind Heradstveit's

Norsk Data—A Success Story; the IEEE for use of Policies ; and the

B.G. Matley and T.A. McDonald's National Computer

University of California Press for the use of extracts from Joseph M.

Between Dependency and Autonomy: India's Experience with the International Computer Industry. I would also like to thank Business Week, Business India, the Deccan Herald, the Economic Times, and The Hindu for permission to use extracts Gricco's

from these various sources, and I.G. Patcl for permission to quote from his Kingslcy Martin Lecture. Secretarial assistance was provided at various points of time by Ms K. Sashikala, MsS.M. Padma, MsT. Kamala, Ms Audrey Alfonso, Ms Arthi Vcnkatachalam, and Ms Vijayalakshmi Bai. Ms Padma Ganapati read the typescript for initial corrections. All of them have helped me immensely in the task, and I would like to record my thanks to them. The publishers, the Oxford University Press, have taken a great interest in the book and I must record my thanks to them for all their help and advice. I would like to affirm that the views contained in the book are entirely my own and do not reflect, in any way, those of the organizations who have so generously funded the work.

A Note on Units of Currency and Value

A 'crorc' equals 10 million. A 'lakh' equals 100,000. In 1966 the exchange rate of the rupee was fixed at Rs. 7.50 to the US $. The rupee was later put on restricted float and the value has been declining; in 1983-4 the exchange rate was about Rs. 10.34 to the US $; in 1987_8, RS. 13.09; and in 1990, Rs. 17.00. In July 1991, the rate declined to about Rs. 26 to a US $ after a steep devaluation. The purchasing power of the rupee has also been declining

ue to

inflation. In relation to I960 prices, the purchasing power in 1967 was 58 1- in 1975 31.2; in 1985 16.4; and in 1988 13.0. The cost figures and prices in the book refer to the then prevailing value and exchange rate.

CHAPTER 1

Self-Reliant Development Fails (1968-78)

EARLY DAYS The first digital computer to arrive in India was the HEC-2M which was specially designed and built by A.D. Booth at Birbcck College, London for the Indian Statistical Institute (ISI) in 1955. 1 P.C. Mahalanobis, a leading economist, one of the main architects of India's Second FiveYcar Plan and Head of the Indian Statistical Institute, Calcutta, initiated the project. The HEC-2M was a valve-based machine with a 1K memory. In 1958, under a grant from the United Nation's Technical Assistance Board, ISI acquired a second computer — the URAL, from the Soviet Union. IBM was to supply a third machine in 1964, an IBM 1401, which was to supplant the other two. The first commercial computer was installed by ESSO Standard East­ ern Inc., at Bombay in 1961. From 1962 to 1964 fourteen more computers were installed, out of which as many as twelve were placed in research and development organizations. T1FR (The Tata Institute of fundamen­ tal Research, Bombay) acquired a large CDC 3600 system in 1965, operating it as a national facility. The pace of computerization quickened from this time, with thirty commercial installations coming into existence in 1965 and 1966. During this period the role of the international companies had not been entirely positive. 1CL (International Computers Ltd., UK) had been operating in India mainly through leasing obsolete unit record equipment. IBM (International Business Machines Ltd., USA) had also entered the Indian Market with mechanical tabulators and accounting machines. It has been noted that over 1000 mechanical systems were in operation in

£ O c


> — k- . a. - •- « w u 3 . 00

3 oo

ob'3

re .5 ~ J= J 2 -5 k- -o O 2 .=

•a 50 5 to 50 4>ort term gains which might pre-empt our long term position in this matter.... 1 do believe that major international companies have a contribution to make in India, on honourable commercially rewarding terms, on ground rules which are acceptable to India. 1 do not think that this could be at the cost of a major national ettort, which we must put out without further delay. This major national etfort will require, first of all, a major information system ... a corporation of some sort that is funded on the basis of investment of some Rs. 20-30 crores in people, ideas and software.(emphasis added)

He was voicing his unqualified support for the Electronic Corporation of India's (ECIL's) computer programme and for a national informatics organization. The Conference ended on a euphoric note with high hopes for a self-reliant electronics industry to be captained by a scientist, with a team of scientists and engineers.

ELECTRONICS COMMISSION AND DOE FORMED The Electronics Commission was set up in February 1970 under the Chairmanship of M.G.K. Mcnon. The DOE was set up on 26 June 1970, with Mcnon as the Secretary of the Department. The Department func­ tioned directly under the PM. It was considered a 'scientific department'.

8

India and the Computer Sarabhai did not live to see his dream fulfilled. He died suddenly on

30 December 1971 in Trivandrum, leaving Menon to shoulder the burden of both the Space Commission and the Electronics Commission till S.Dhawan took over the Space Commission. With these appointments the domination of the regulation, licensing, control, and development of the electronics industry and its technology by scientists commenced. The computer industry and its technology became subject to 'scientific' treat­ ment. M.G.K. Menon had by then mentally prepared himself to leave TIFR and to enter government. In various public pronouncements he had advocated the idea of scientists entering government. He wrote in 1969: 1 6 It is my overall impression that the scientific community, as a whole, feels itself today to be out on the periphery. It has little to do with policy formulation and decision making, certainly not on broad aspects of national policy, and far too little even on aspects which have high scientific and technological content. ... On the other hand, if one could convey properly to the scientific community the enormous tasks which the nation faces, a great many of them with high scientific and technological content, and could get the scientific community excited about the possibility of participation in this venture, then one would indeed have achieved a breakthrough. ... For all this to happen, it is vital that the scientific community is brought into the picture in a much more vital manner than at present.

... As a first step, the administration of science should be placed in the hands of scientists,(emphasis added) After having been appointed as Chairman, Electronics Commission and Secretary, Department of Electronics, Menon had the task of laying down policies for the development of the whole field of electronics, including the area of computers. In charge of the computer desk at the DOE was Col .A. Balasubramanian (who later rose to the rank of MajorGeneral, while serving the Defence Research and Development Or­ ganization). N. Scshagiri provided inputs for planning and policy making, being in charge of the Information, Planning and Analysis Group (IPAG) of the Electronics Commission. In the wings were R. Narasimhan and P.V.S. Rao of the (TIFR). Rao also served as the Chairman of the Technology Development Council Working Group concerned with com­ puters. Inducted as the Secretary, Electronics Commission (also titled, Ex-Officio Joint Secretary), Ashok Parthasarathi assumed the role of policy conscicncc-kccpcr. lie had earlier been chosen by Sarabhai and A.S. Rao to present a paper at the National Electronics Conference which sug­ gested the setting up of an apex body — the National Electronics Board

Self-Reliant Development Fails

9

___ which later evolved into the Electronics Commission and the Depart­ ment of Electronics. A.S. Rao, Managing Director of EC1L (Electronics Corporation of India Ltd.), was a member of the Commission but his views were considered partisan in respect of computer polices as he championed the cause of ECIL. He wanted policies to be oriented to making ECIL the 'national champion' in the field of computers. H.N. Sethna, Chairman of the Atomic Energy Commission (AEC), and also thc

Chairman of ECIL, prompted by A.S. Rao, challenged Mcnon's

initiatives to open up thc industry to other entrepreneurs. Given Mcnon's introspective attitude and penchant for debate and discussion, and his very studious nature, policy formulation was long drawn and decisions took time. There were also difficulties in putting into action thc concept of self-reliant computer technology development and manufacture. Faced with thc formidable task of implementing this ambitious scheme within a five year time frame, the Department of Electronics chose to dub thc approach of thc Bhabha Committee as 'tentative and preliminary' — which proved to be a way out of this daunting task.

AN INTERNATIONAL PERSPECTIVE Indian scientists arc a well-travelled lot and many obtain their post­ graduate degrees in foreign universities, going on to hold research posi­ tions. They were also well informed about international trends and the developments, with foreign technical journals and magazines at hand to provide the latest information. It is, therefore, logical to expect that thc members of thc Bhabha Committee would have made full use of such information in making rational and pragmatic plans for computerization and thc development of thc computer industry in India. It may be noted that in the US thc total market (see Table 1.3) for computer and com­ puter-related equipment and process control had already reached a figure of about US $ 2 billion by 1965 and it was expected to touch US $3.4 billion by 1969. This growth had involved enormous investment in research and development over a wide spectrum of electronic com­ ponents, and large expenditures in the evolution of computer languages and software. It may also be pertinent to note that the total federal funds made available for Research and Development in thc US in 1966 amounted to about US $ 15.32 billion, apportioned roughly as shown in Table 1.4. An industry-wise breakdown of this expenditure is available and

t co O u T NO >/-> o o o o o d o — vi oo « 00 o q o ri «o o>" o r-_ o — 0 fs r> 0 0 0 0 0 0 -0 n >0' n pi «

3 OU«£ vO vo o

0 0 0 ~ >o o •— o o

^
o pi >ri

o c o

q 0 >0 n fO g

0 q 0 (N n r~ o o o g ^ pi

LU 2 2 2

5> 25w !cc 0E 0 t> 0 s 2 Eu COE E •O 2 E 8. E3wu 2 T3 8 . 3 60 20 0 u t: c e (/) ca> icM > 4 -o VIS.1 00 re CO CO o o U CJ oc oc 2 2 2

.2 •a P 8 S? 2

_ e EJ2 o-

2 E .2 c 3S S 2

a

Self-Reliant Development Fails Table 1.4 US Federal Funds for Research and Development (1966) (US $ billion) Defence

7.02

Atomic Energy

1.21

NASA

5.05

Others

2.04 15.32

Total

US National Science Foundation, Federal Funds for Research, Development and Other Scientific Activities, vol. xxi. Historical Statistics of the United States, Colonial Times to 1970, part 2. Productivity and Technological SOURCE:

Development, Series W126-43. Sourced froml947-51, US Bureau of the Budget (now the Office of Management and Budget), unpublished data: 1952-70. indicates that electrical equipment and communications absorbed over US $ 3.62 billion. The early developments of office machines, computers and components for computers received a very generous portions of this outlay (sec Table 1.5). This annual level of expenditure was a sustained effort, and in the later years, tended to increase. Table 1.5 R&D Expenditure in Industry Office Computing and Accounting Machines Year

1965 1966 1967 1968

Total Expenditure (US $ million) 549 573 614 678

SOURCE: National Science Foundation, Research and Development in Industry, 1981 (NSF 83-325). Table no. 86, 1965; Table no. 58,1966; Table no. 56,1967, Table no. B-42,1968. This expenditure rose to about US $ 1.765 billion in 1975 and reached about US $ 1.971 billion in 1981.18 In the US, as in most western countries, development work takes place in parallel and on a competitive basis, which is perhaps wasteful of resources. On the other hand, this

12

India and the Computer

allows alternative avenues to be explored, and the most elegant economic solution survives. In the case of the early IBM 360s, hybrid microcircuits (because they were faster) were used, whereas other US computer com­ panies chose the monolithic integrated circuits. It could be argued that development costs in India would be a fraction of those in the USA because of lower manpower costs, but decision making in India is pain­ fully slow and the procedures involved consume an enormous amount of time, nullifying the low cost advantage of engineering and scientific talent. Emerson W. Pugh, in his book, Memories that Shaped an Industry,19 has given an exhaustive account of the development of the fcrrite core memories of computers, and especially that of the IBM system 360. He notes that IBM finally agreed in 1964 to pay MIT a sum of US $15,784,000 for the use of the basic Forrester patent, and that IBM alone spent over US $ 26 million to develop commercial memory products. IBM held seven patents of its own engineers. In addition, it obtained, under licence, five other patents. The development of fcrrite core memories spanned fifteen years and the cost, per bit, dramatically decreased with improved production technologies. They have now been superseded by semiconductors. To provide India with computing power, when billions of US dollars had already been expended on research and development, and when sums were continuing to be spent at a rate approaching billion dollars a year, the AEC/Bhabha school of scientists who had assumed power wanted a self-reliant development approach where all the computer systems and their components were to be indigenously developed and manufactured. Development was to proceed in parallel. The approach was neither pragmatic nor realistic. Technology was to be bought only for peripherals and for some components. Foreign companies were not to be allowed to operate in the country cither. The report of the Electronics Committee did not wisely forecast the financial outlays on design and development that would be involved in putting this plan into operation. The Electronics Commission and the DOE were left to grapple with the matter. This report of the Electronics Committee, spelling out self-reliant development in all fields of electronics including computers, and the later affirmation of the same views at the National Conference, placed the Electronics Commission in a tight corner. M.G.K. Mcnon, as the first Chairman of the Electronics Commission and Secretary, DOE, had to accept the approach — he was, anyway, a part of the TIFR/AEC community. He became virtually a prisoner of the concept of self-reliance

Self-Reliant Development Fails and selective computerization, inhibiting natural market development. The difficulties that confronted him in the execution of the policy forced him later to change his outlook, but, he started his career as a scientist/ad­ ministrator with the lofty declaration that the electronics industry in India will have to give up 'soft options' and start production based upon indigenous design and development. Though technology import was permissible according to the Electronics Committee Report, efforts in this direction were not looked upon with favour. M.G.K. Mcnon had another brief — to reorient the administrative set up so that the aspirations and activities of the scientists would not be curbed by the bureaucracy. He was also expected to provide the speedy administrative support then already being provided by the Atomic Energy Commission for the Atomic Energy programmes.

POLITICAL OVERTONES Mcnon's problems were compounded because he had also to concern himself with the prevailing political attitudes towards computerization. The Ministry of Labour, Employment and Rehabilitation had set up a high-level committee in 1969, initially under the chairmanship of R. Vcnkataraman, (Member, Planning Commission at that time, later the President of India) to look into the matter and advise the government. The resolution setting up the committee had noted: Government's policy in this regard has been that automation could be introduced on a selective basis and that technological advance should be regulated to make it consistent with the good of the community. The criteria which determine such selectivity need to be more clearly defined, and for this purpose the Government have decided to set up a committee.

V.M. Dandckar later took over the Chairmanship of this Committee. The Automation Committee favoured a positive approach and recognized the legitimate use of computers in the fields of education, science, defence, and even in some commercial and industrial establishments. But the strict regulatory measures they prescribed were contrary to their positive approach. They wanted all proposals to be scrutinized by two experts, case by case, and a 'justification report' furnished. They preferred computer facilities to be set up in the public sector as central facilities. They also made prior agreement with the workers mandatory before computers were inducted into any organization. Satish Loomba, a Labour member of the Committee, in an explanatory note had added

14

India and the Computer

that while in a socialist system automation was beneficial to the people, in a capitalist system automation would lead to increased exploitation of workers without leading to any social benefit, adding: 'India is not yet a socialist society'. The big business representatives — Naval H. Tata, Babubhai M. Chinai and B.D. Somani — found the procedures being laid down in the report onerous. They noted that the computer is a 'neutral agent, and its advantages are evident in all economic structures, whether developed or developing, whether centrally controlled or mixed'. The Administrative Reforms Commission had also their say in the matter. In their review of LIC, they had noted: 2 0 In view of the recommendations made by us already regarding decentralization of the powers and functions of the Corporation to the Branch Offices which will in effect, be functioning like satellite insurance companies, we do not consider that computerization is an urgent necessity. Under our scheme of reorganization the Divisional Offices will have to make full utilization of the Punch Card Machinery already in use at these offices. Mcnon had also to face H.N. Mukcrjce, Chairman of the Public Ac­ counts Committee, and explain his stand. 2 1 In a note furnished to the Committee, the DOE had stated: The majority of the computers currently in use in the country are largely data processing machines which support essentially accounting and other book-keep­ ing functions. This was the result of the way in which the computer industry has been promoted by the foreign controlled companies. Mcnon further stated: ... that the approach we have to take in this country has to be predicated by the fact that we have abundant manpower and, therefore, we have to use them in the most effective way ... this must be the basic philosophy with which we should approach the problem. Mis cautious attitude was appreciated, and the PAC noted: 2 2 The Committee find that the DOE has so far been designated as the nodal point for scrutiny only in the respect of government departments. The Committee stress that all requests for introduction of computers above a certain value, whether in the public or private sector, should be . . . scrutinized by the DOE which has the requisite expertise to make sure that such computerization would subserve the larger public interest. The PAC concluded:"" 1

Self-Reliant Development Fails While the Committee arc sensible that our country must lake every possible advantage of modern scientific advance and technological innovation, they would stress that in an economy such as ours, where the problem of unemployment is large scale and of serious proportions, the use of computers and other sophisti­ cated machines for traditional labour-saving applications may not be desirable or even expedient. The common argument that computers lead to efficiency, which in turn, leads to profits, larger savings and faster economic growth is, perhaps, valid only from a long term point of view. The use of computers for predominantly labour-saving applications can well result in contracting the employment oppor­ tunities It is unfortunate that adequate attention has not been paid to this impor­ tant aspect in the past. The Committee desire that the government should in­ variably take into account the social cost of computerization and evolve a prin­ cipled and positive approach for the selection of areas for computerization on an over-all judgment in the national interest.

Mcnon had also set up a think-tank to assist him. The Working Group of the Technology Development Council (TDC) headed by

pV-s- Rao

had studied the problem of computerization and their report,

though

supportive of computerization, prescribed strict scrutiny and regulation for the introduction of computers into the Indian environment. The Automation Committee Report, the scrutiny by a Parliamentary Committee on Computerization in government departments and the TDC Working Group recommendations, all made Mcnon wary, as they showed computerization had political and social overtones.

MENON INITIATES PROGRAMMES AND UNDERLINES SELF-RELIANCE Mcnon proceeded in line with the sentiments expressed at the National Conference and as outlined in the Bhabha Committee Report. He served notice to IBM and ICL to conform to FERA (see Chapter VII), to shed their trading activities and switch to the manufacture of their latest products or to operate as 100 per cent export-oriented companies. He initiated steps to set up regional computer centres and persuaded the government to approve the formation of Computer Maintenance Cor­ poration (CMC) to take over the maintenance of the IBM and other imported systems as there was every chance that IBM may not accept the conditions being imposed by the government. A new measure — the formation of the National Centre for Software Development and Com­ puting Techniques (NCSDCT) — was instituted with the help of UNDP. Mcnon obtained UNDP assistance for the National Informatics Centre

16

India and the Computer

(NIC) and he approved financial grants for the Bharat Electronics Ltd. (BEL) peripherals development programme. Piecemeal permissions were given to ICL to assemble and market their ICL 1901A systems. BEL was to be the subcontractor which assembled and tested the systems for ICL. Software export promotion steps were initiated. We shall look at the relative success of these measures in detail later. In addition grants and loans were given to ECIL for the development of the TDC scries of computers and associated software. Other technology development schemes to be initiated included an important project for the Defence Ministry. BEL had wanted to work on this project using imported knowhow but Sarabhai had vigorously fought the issue, leaving Mcnon to fund TIFR for the project. This gave TIFR a foothold in the Defence Ministry and the Electronics Commission a say in their affairs. In the event, the TIFR project was delayed, leaving the Defence Ministry a wait of nearly a decade for the necessary items. In order to govern both computerization and the import of computers, the Electronics Commission took the approval of the Cabinet on 4 December 1975 for strict procedures. The procedures (see Appendix B) remained in force till liberalization was found necessary in 1982-3. It was decided that all users should first attempt to meet their in-housc requirements through those computers which were available in the Indian market or through use of the regional computer centres. However, the setting up of these centres was a slow affair, taking five or six years to become operational. Computers other than those available in the Indian market were to be regarded as expensive items of import. The regulatory role of the DOE concerning computer imports came into full play, and the zeal with which this control was exercised was to cause serious delays, and later bring discredit to the DOE. Case histories of computer imports compiled by the Review Committee on Electronics (Sondhi Committee) reveal the serious delays. 2 5 Time estimates in the procedure were unrealistic and serious hold-ups were encountered by both the public sector and private sector applicants, and by several government departments and R&D institutions. Many considered the delays intentional to give ECIL a chance to meet the market demand. The 1972-3 Annual Report of the DOE mentioned: 2 6 There is no question that the area of computers is going to be of great importance for national development and growth of electronics; it must be regarded in the widest sense as an area of digital electronics and controls. From the point of view of its growth potential and strategic importance, the Department of Electronics has taken the stand that indigenous capability and self-reliance must be generated in

Self-Reliant Development Fails this area as rapidly as possible. The policy principles for the computer area were outlined in the 1973-4 Annual Report of the DOE and ^following criteria for planning the computer industry were laid down: a) The requirements for large computers will be essentially me

y

imports during the Fifth Five Year Plan period. b) The programme would aim at self-sufficiency with regard to re­ quirements of CPUs for the minicomputers and medium computers by the end of the Fifth Plan period. c) The programme would aim at rapidly building up a self-sufficien industry with regard to a major part of the requirements of peripherals, including import of know-how where required. . d) The requirements of software would essentially be met by an indigenous base which would be actively built up with the further aim

of generating an export potential. c)

The country would be self-sufficient with regard to capability to

maintain installed computer installations. , . The Estimates Committee of the Lok Sabha which reviewed the Department of Electronics were also informed of the policies of the DOE: 2 8 The Committee note that it is anticipated that the country will be self-sufficient in respect of requirements for electronic calculators and minicomputers and a so with regard to nearly 90 per cent of the medium sized computers by the end of Fifth Plan. The Committee also note that the requirements in respect of large sized computers and some sophisticated peripherals and components will continue to be imported even after the Fifth Plan. It is in the field of policy formulation for minicomputer production that Mcnon faced difficulties and came into conflict with A.S. Rao, Managing Director of EC1L, and H.N. Scthna, Chairman of the Atomic Energy Commission and Chairman of ECIL. He could only finalize a policy just prior to his being moved from the posts of Secretary, Depar ment of Electronics and Chairman, Electronics Commission.

THE MINICOMPUTER PANEL Though ECIL was to be the leader in the area of computer manufacture, Mcnon was under pressure to let others from the private sector take up minicomputer manufacture, quite a few of whom were entrepreneurs willing to exploit the new, evolving technologies. He set up a panel in

18

India and the Computer

June 1972 under P.V.S. Rao of TIFR to survey the field of minicomputers and make recommendations for the development of the technology and the industry. IPAG published the preliminary working document.

The

panel estimated the requirements for minicomputers for the Fifth Plan period, 1974-9 (sec Table 1.6) and compared their assessments with those made earlier. The panel recommended that licences should be issued to twelve applicants, each with a capacity of fifty to sixty machines per annum. The view was that at least six to eight units would reach full production. The Committee noted: In the area of system design, software, logical design, mechanical design ... and so on, the level of technology and know-how available within the country are such that minicomputer mainframe manufacture can take place in India without any know-how being imported from abroad. The local industry and the R&D estab­ lishments are fully capable of ensuring a satisfactory minicomputer industry of the highest order without any external help.... The minicomputer industry can be established and maintained as a wholly indigenous industry.... No import of know-how is necessary for the establishment of a minicomputer industry. No foreign participation need be allowed in the field.(emphasis added) The standardization of designs, though difficult, was considered an essential prerequisite to local manufacture in view of the variety of LSI chips available, there being no prospect of the local manufacture of these devices in India. The Panel also felt that there may be advantages in bringing into existence a publ ic sector corporation exclusively for computers and allied products. They were hinting that the ECIL computer activity should be separated and placed directly under the DOE. A similar view had been expressed in 1973 by P. Jayant, CSI, who had written to the IPAG on this matter. 3 0

OBJECTIONS AND DELAYS The idea of throwing open the minicomputer industry to several entrepreneurs did not find favour with ECIL. S. Srikantan of ECIL wrote to the IPAG Journal :31 In order that we do not waste our time, talent and material resources, we must pool our efforts initially into a single unified project of sufficient size, what has been referred to as critical size, that will sustain self-contained growth. We must remember that the priorities of computer application should be decided in terms of real local needs and computer development should take into account these priorities; we should consciously desist from adopting a path in mere imitation of

— .2 c O X! (/> ._ —• u s —— p p t E »1 « J - E ", £"b« _ '5 o 2.?

Ji •S

£Vl > § o

.i" — X

S5 >•;.!> g Q — w > cL

io o o cs

o o oo o QvOO"!^ CM co

o©

>o o >o »— -«t w

•eoMON co r- v to >o tM to —

v> o 00 o

\o o o oo o CM

C7> o o 0"tN

CO ©

CM

C) >

£ -C

2 f s©

2S8

S S S g

^*>2

S2S2

^22222§

•t >n ru

moan — ~ —

lOioofsm-rou— in —

— CN IN

|2 2

f^io-OfNmTforr

n

2

, T

°2

S

£ in oo

W

~o 2

iJ

t/i B -i §• ~ 5.g "S o c3

^ :§

Z ^5

m

lis

E •= 2 *1 °°,

o

Z

9

in o-'

o o m m Tt

r—

88 a.

oT

ma

O

•5

>

>

E

a

00

E

oo e

o

z

E

c

>

,A

w

>•

*3- a-

•S c w >

tn

£

8. £ k> W

00

TJ

C

£

E

2

co «

O H 0 in

1

ri

£

«5

I

H

g

a. O

00 *t m t-~ •— i-

o

rn r~

oo — m

3 Q.

0 oc c

12 w

I* E

z

S2 o

5 o

1

52

-c

NO ri _w JO !3 f-

tf £ a. c

CJ o

I

0=

•o c

a

1 I UJ

E o Q

o

Z

— a.

•s

uj

a

j=

C/i

c o

2 0-

o

c V > CJ

I

E o U u. o

^ _

'C

"5.

L/L I-

O

» Q. Q. 00 ^ O

c/i —

.E ra

C

—

!j

5

>

u.

2

S

x

^

x

98

India and the Computer Table 4.6 Summary of Demand Projections - Peripherals (Rs. m i l l i o n ) Normal Optimistic

Seventh Plan

Eighth Plan

Period

1985-90

6,000

8,500

Terminal Year

1989-90

2,000

2,800

Period

1990-5

11,000

15,000

Terminal Year

1994-5

3,500

5,000

SOURCE: 'Report of the Study Team on Computers/Office Equipment including Software', IPAG Journal, vol. 12, 6, New Delhi, March 1985, p. 310.

The Col. Rao Committee made the following recommendations:24 1.

Considering the economic viability, scale of production for effec­

tive absorption of technology and in-depth indigenization of parts, it has been recommended that by 1989—90, the effective and functioning units for each of the major peripherals should be as listed (see Table 4.7). 2. As far as possible, India should go in for central purchase of technology and encourage equity participation from the foreign col­ laborators so as to keep abreast with technology. 3. There is a need to promote units for the manufacture of print heads, read/write heads, servo motors, etc., for OEM supply to peripheral manufacture. 4.

The purpose of canalization should be clearly defined and the

canalizing agency should be provided with appropriate guidelines which needs to be followed. The canalization could offer the following economies: (i) Canalization could secure economy of scale of purchase, by bulking of imports, thereby making available to the Indian computer user, peripherals at prices much lower than what he would have had to pay through multiple imports from multiple vendors. (ii) Achieve standardization of the models in respect of each type of peripherals, as a prelude to the setting up of its economic indigenous manufacture. (iii)

Provide the necessary maintenance and support, economically,

Peripherals Pose Difficulties

99

Table 4.7 Peripherals - Licensing Recommendations - Rao Committee Units

Item

Dot matrix printers

Annual/ Capacity/No.

5

5,000-10,000

10

5,000-10,000

VDUs (including graphic terminals) Floppy drives

5

10,066-15,000

Tape drives

3-4

1,000

Line printers

3-4

1,000

Disk drives

3-4

2,000

SOURCE: 'Report of the Study Team on Computers/Office Equipment includ­ ing Software', IPAG Journal, vol. 12, 6, New Delhi, March 1985, p. 308.

so that the reliability, and hence the availability of the computer industry is not adversely affected. 5. In view of the high stakes involved in the stable availability of the complex peripherals, there is a need to promote a public sector unit to take up the manufacture of floppy drives (15,000 per annum), disk drives (2000 per annum), tape drives (1000 per annum), and line printers (1000 per annum). Induction of a public sector unit would help control prices and could provide the initiative for R&D to ensure technological abreastness of these peripherals. The Study Team also noted:25 In the Seventh Plan period, it is proposed to establish a centre for the development and production of computer peripherals. The total cost of this project is estimated as about Rs. 200 million during the Seventh Plan period. The impact of this project would be helpful in the attainment of self-reliance.

However, the establishment of a centre for development and produc­ tion of peripherals did not find a place in the DOE list of Plan Projects for the Seventh Plan though the target for peripherals production was fixed at Rs. 200 crores.

NOVEMBER 1984 POLICY The November 1984 policy provided for the liberal import of know-how and promised that the excise duties would be kept at a low level. The

100

India and the Computer

duties on peripherals not made in India were reduced to 25 per cent, and the duties on the other peripherals were also reduced. Those on mechani­ cal parts for peripherals to be made in India were reduced to a new low of 5 per cent to facilitate the establishment of a SKD/CKD-based com­ puter peripherals industry. The electronics policy announced on 21 March 1985, broad-banded the peripherals area. MRTP companies and 40 per cent foreign (equitybased) companies could now take up peripherals production. The aim was to make computerization less expensive. But a consistent policy did not result. The duty structure, as well as what could be imported under OGL underwent many changes, taking away the initial liberal thrust.

A Growing Interest in Peripherals Production As the market for computers picked up, with the sale of PCs alone destined to reach a figure of 64,000 in 1988, there was a greater interest in the production of peripherals. The reduction of the duty on mechanical parts to 5 per cent (enhanced in the 1987 budget to 98 per cent, again reduced to 50 per cent on representation, and now refixed at about 84 per cent) also encouraged traders and other parties interested in SKD/CKD operations to seek an entry. A spate of registrations occurred after the industry was dcliccnsed, but most of the parties failed to pursue their projects after registration. Table 4.8 summarizes the licensing status for the major peripherals. The table docs not include the many parties who had merely registered. That standardization was no longer an article of faith of the DOE is obvious from the manner in which the industry was sought to be promoted. Neither did the issue of the economies of scale of production bother the DOE — this was now entirely a matter for the entrepreneur to decide. Apart from several small parties, the major companies who had come forward included Larscn & Toubro, Godrcj, Wipro, Philips, and the TVS Group. Splintering the market among so many parties could not bring about any substantial reduction in prices on account of increased competi­ tion. On the other hand, none of the companies can have a great enough turnover so as to be able to afford R&D, leaving CKD/SKD methods of production the norm. Models arc bound to change in the international market making it impossible to promote a genuine industry where there is a value addition of at least 60 per cent. Peripherals like dot matrix printers are now more electronic than mechanical. Except for the print-head, the other mechanical and plastic parts arc well within the capability of the Indian industry. VLSI and LSI

D U0 1 5

x

XI

E

3

z

I —

w Z

l»J

•2 _ O u-

re . o —• o u

re

E 2 9> © O

c

GO >

re 3 U Z

g.E

O O O in *-"„CJ

So

2 o CJ

£ s

«—
a.

o o

•o c

in

CM

o m

oo oo *9" CM

& r-« 8J

SO TT Tt VO

I a Q

CM O O O 00 O in O CM vo •— d •—• o cm rO CM

£ CL £ O Q

(/)

a.

E

o U

!

— O tu os W . . CL > ~

o o o ^ w [/) 'e ;C * o O >» .2 £0 Cfl o. J: •- "O t> -> M C c O — ^ 2 u « « ca h "" 2 r _ Q « w 0 lv -w O. jT ~ c OJ^_-O602.CC4) 03 — • -h "a '? 10 •— ,2 70 .E ? ;= W « C C_2 • —»• -• - —,iD a

. HKv/ril Honevwell-Bull, it was felt that a second

digenously was taken mainly duei to ^^"^n'clearance to

^^X^r^arin

^A.omieBne,

gy, Space, etc.

cent interest-

Q0

At the end of this penod, the actm^

ject cost was of

imatcd

French training of Indian engineers a

sjsrrsrsa-^r-?!s,.million.

estimated cost of Rs. 45 lakhs. ^

CII-Honeywell-Bull

February 1985: As the evaluation of the tendersi forth.'

3"°'°hfs L, been set for the

As Dataquest noted in February, 1985: Some financial newspapers have been has gone in favour of Bull o

ranee

Control Data ^dismissed the publish-

taken as yet.

n

A correspondent of Dataquest noted in July 1985: Do we really have «),,aboral;c^ ^^rrtK'^ion. -P-"* The Dc°

the

Hindu, Madras, reported from Washington

CD°" COTSiderCd h"e^ a ""ol-in-the-arm for the troubled Amp116'""" troubled American company.... However, the Washington Post observed 'while considered a powerful machine, these are Control Dala's entry levd mainframe

^0!^' ''r,3601" f°Ur yEarS °W a"d

USe comPuler technology that is 10 years old. They will „„„e ,he less be ,he mos, powerful Indian-made^mpuieis

forTCDC:™e """"h

Da"">UeS' 1UO!cd

™g^i"e on the outlook

:0mpa"),'S busi"e» 18 in deep red. ^r.Td'r b'erkClal r*™"*8 subsidiary. It is in default on nearly US J 400 milr""u rly US S 400 million .... Hence ,t is more than evident that the writing is on

Mainframes - A Late Awakening

121

the wall for Control Data. But what is intriguing is that the decision on the collaborator is being made possibly without taking into account the long term ramifications of CDC going under.

CDC Agreement: ECIL's New Problem The formal agreement with the US was signed in September 1986. Peter McPhcrson, USAID Administrator, signed the agreement on behalf of US, and Ambassador P.K. Kaul on behalf of India. A US $ 7 million grant under USAID and a loan of US $ 20 million from the Export-Import (EXIM) Bank at an interest rate of 7.4 per cent for a 10 year period cemented the deal. The assistance package was one of the twelve socalled 'mixed credits' assembled by the EXIM Bank to help US com­ panies compete against foreign firms similarly subsidized. The Technical Evaluation Committee, after two years of deliberation and several visits abroad, had stated that the CDC proposal should be rejected. That CDC was finally able to obtain this collaboration was probably because of the keen desire of the Defence R&D Organization to obtain technical co-operation with the US. The commercial prospects of the project were ignored. Yet ECIL will find it hard to sell these high priced computers to thc-railways or banks, who will probably opt for more economic solutions for their operations. Neither can defence research absorb the 30 or 40 computers a year ECIL will produce. It is a pity that peripherals were not covered in the technology transfer agreement. The cost of peripherals runs high, and imports are un­ avoidable. Though CDC's financial position had deteriorated, their stand­ ing as a technology leader had not been in doubt, and India could have simultaneously found a solution for peripherals production by tying up with CDC, thereby forcing standardization, at least of the high-end peripherals, from a single indigenous source. With Bull and ICIM enter­ ing the mainframe market, peripherals arc going to be imported from various sources, leading to a continued lack of standardization. This situation could have been avoided with a little forethought. It is now too late for this to be arranged, as in 1989 CDC was forced to sell its peripherals activity to ScaGatc Technologies because of an acute cash crunch. It had earlier closed its ETA supercomputer production for similar reasons. Even after the agreement was signed, Scshagiri had doubts about the wisdom of going ahead with the manufacture of the CDC 830 system. He is reported to have told the Electronics Commission in December 1986 that the Super 80 computer had much more power than the Cyber 830,

122

India and the Computer

and because of this, the DOE was in a fix whether to go ahead with the manufacture of the Cyber 830 system or drop it altogether. Deodhar, the Chairman of the DOE, is reported to have felt that after a thorough study and evaluation, they could go back to the Cabinet and state the exact position. That the CDC 830 series was not what the country wanted has been made amply clear by the fact that the imported systems have been given to the universities by the Defence Ministry — the Scientific Adviser generously donating them in the hope that computer scientists trained at these centres would later help the Defence Research effort. The recipient universities/institutions were: Cochin, Vizag, Mysore, Madurai Kamaraj, Indorc, JK Institute, Allahabad, and Poona. These systems were not exactly what the universities wanted, however. A scheme had been formulated earlier, a Committee recommending in 1984 the VAX 8350 as the best choice for the universities. No progress had been made in implementing the scheme, so eight of the CDC 830 were disposed of in this way. The DOE/ECIL mainframe project did not target the univer­ sities amongst its first priority recipients. The DRDO also allocated one CDC 830 to the Naval and Occanographic Laboratory at Cochin and another to the Instrument Research Dc.clopmcnt Establishment at Dchra Dun. Another 830 went to the DOE for the CAD/CAM development programmes, and a further one was taken by the ISRO. Even after the initial allocation, many of the universities could not receive or install these computers because their facilities were not ready. Quite a few of the computers remained unused for more than two years — surely a fit case for an audit to look into. Faculty resources appear limited in many of these universities and will have to be augmented for an effective use of the computers. For CDC, the order of fifteen 830s, together with their peripherals and spares amounted to a cash injection of about US $ 10 million in respect of hardware supply alone. They will also now get an order for seven 930 systems (because of a change of model). In addition, know-how fees (US $ 4.5 million), software (US $ 5 million), and spare parts sales will boost their income. Royalties will only arise if, and when, ECILstarts manufac­ ture, and even then, parts supplied by CDC will be excluded as per the usual Government of India norms. Nevertheless, CDC Indo-Asia Com­ pany could indeed feel satisfied. DCM Dataproducts also has cause to be happy. The agreement with the Government of India did not shut them out. Loopholes existed in the agreement with CDC which permitted DCM to sell CDC computers,

Mainframes — A Late Awakening

123

other than 810/830, in India. This undermined the ECIL effort. CDC 840s have been sold by DCM to IITs (Powai and Kharagpur) and to the RCC at Calcutta. The CAG has adversely commented on this DOE-funded purchase, noting that the site in which to install the computer had not been decided upon, and that the delays incurred cost the DOE a further sum of Rs. 142 lakhs towards custom duties, etc.19 These institutions could just as well have bought the 830 through ECIL or waited for the 930. The Scientific Adviser to the Defence Minister could be well pleased. He had got what he wanted. The IBM 3090 (made mostly in Japan and sold through IBM, Australia) arrived, and has been installed at the Bangalore DRDO Computer Centre (without the vector processor which the US Government refused to clear) with the Catia aerodynamic design software and a number of workstations. Some declassified software packages, such as 'Cosmic' for aerodynamic design, have also come in, courtesy NASA, USA. Integrated Computer-Aided Engineering and Manufacture software packages used by Northrop and others have also arrived. The Defence R&D scientists now have enough computing and software power at hand to help solve their problems and to get the LCA flying. The concentration of computing power within one square mile of the HAL airport at Bangalore is immense — but the systems are diverse and portability of the programmes, even in Fortran, is not practicable. NAL's computational fluid dynamics packages developed on an old Univac cannot be run on the IBM machine. Currently, the utilization of the immense computing power appears low, with several PC ATs reduc­ ing the demand on the mainframes. The realization dawned late upon Electronics Commission that they should start the manufacture and design of mainframes in India. It took the Commission, the DOE, and the Government some 5 years to take a decision on the collaboration, making the views of the Technical Evalua­ tion Committee irrelevant in the process of reaching a decision. The choice of the CDC, linked with the attempt at softening the attitude of the US Administration in allowing the import of other supercomputers and defence-related hardware, is at best, a long shot. It remains to be seen whether this will really help India; the policies of the US Administration arc highly volatile. From the US point of view, the deal helped CDC, a sick company, with an injection of cash, thereby aiding its survival. In the process, it is unfortunate that ECIL will find it difficult to serve the commercial market. Politicians will continue to blame the public sector for poor financial performance. ECIL's current orders for the 'Mcdha' mainframes arc all from the government and the public sector

and

124

India and the Computer

these orders are subject to budget allocations which are somewhat in­ definite at the present time. In summary, fifteen CDC 830 Computers have been imported, no manufacture of the 830 series has taken place; the new model 930/932 has now been taken up for assembly. Screwdriver technology is to the fore at ECIL. In view of the limitations of the CDC 830, negotiations for the CDC Cyber 930 scries have already been concluded. The model 930-31 is rated to be twice as powerful as the 180/830 with a dual CPU. The assembly operations have now started. There is also an on-going dialogue for the 932, the dual processor version, but clearances from the US Government have still to be obtained. The mechanical design of the 932s may be something of an improvement over the 930's and the hope is that even if the 932 is cleared in the single processor version, it will obviate ECIL from again ending up with an outdated model. It remains to be seen whether the Defence Ministry or other government departments will come up with adequate orders so as to make the ECIL/DOE venture a real manufacturing success. ECIL will have to make a determined marketing effort to introduce this computer into the railways and banks. It already appears the banks have accepted the 930 for large branches and the railways for their reservation systems — CMC has installed one for the Sccundcrabad reservation system. But ECIL's programme has already been considerably delayed, though this is no fault of ECIL. ECIL could still benefit from the deal with CDC as it brings to ECIL much needed production technology and current engineering practices. The phased manufacturing programme could result in considerable value addition if the third stage is ever reached. This would probably add about 60 per cent in the case of the CPU and peripheral controllers. But the mainframe area has now been thrown open to two other manufacturers and an SKD/CKD industry is in the offing. Bull have been allowed to tie up with the Bangalore firm, PSI, for the manufacture of their mainframes. The Bull mainframe may prove more versatile and suitable for a wider range of applications, especially for commercial work, where the bulk of the Indian market lies. Bull has taken up a strong equity position in the PSI joint venture so as to ensure management control. And CMC is prepared, and ever ready, to import systems as it is good and easy business. From an economic point view, it is just as well that ECIL is merely executing the project on behalf of the DOE. And what of the future? The scientific community talks about the assimilation of technology, and that technology should not be repeatedly

125

Mainframes - A Late Awakening

imported. Will the next mainframe ctes.gn.com. from EClL?WnMhe throwing open of the mainframe area to all and sundry and a. CKD/SKD throwing p ^ same prob,ems „ dld Wlth ,he TDC STuc compSLn is going to be difficult - the private sector ,s not going to invest in the massive manner of ECIL. They are gorng to rmport

fro'^

out a "cw StratCgy f°FC°mPUtCr

Division.

SUPERCOMPUTERS

tions imposed by the US Government tor use of first Cray has already arrived and has been installed in Delhi tor use oy

™«o £finaiizcd US hesitation. The Institute is making

prcparatto

b;x °f

s;

h;

ur

la

s for t^he tn^allat

.IT Kanpur, were also allowed by the government suncrcomoutcr. ETA Division was a subsidiary of CDC and dum the agents. Now that CDC has closed the ETA division, it remains o

n'

126

India and the Computer

which, if converted to money, could run into a few crores of rupees and lead to another green revolution. Though the possession of a '

fee monsoon predictions could be done without a supercomputer -ail

We have to deeide on where ... to star,... when we have a supercomputer.

AN INDIAN PROJECT: C-DAC The new breed of Indian technology leaders like Sam Pitroda seem to

all the attributes necessary to win the heart and mind of the then Prim!

^:r,'rh-:°8y

"eo

instant

—s'-

a"Crna"VC

'°lhc

whC„ „e

sst

suP«compu,er, approval

was

and'oTrf P°PU'arly rCfcrrcd 10 as ,hc 'supercomputer project', there was and perhaps con.tnucs to be, confusion about the aims of tlte 1988 i"? h frC

PMers Tody 1" February

10 faC° mCC"'"g With Com

Mainframes - A Late Awakening

127

What supercomputer? — complaining that the press has continually grabbed hold of the wrong end of the stick ... neither he nor Sam Pitroda of C-DOT ever had any intention of producing a supercomputer. Theirs were parallel processing projects. But the press liked the idea of a supercomputer, so a supercomputer it became. 'In my view there are no supercomputing projects. It is a parallel processing project, which in my view will bring you up to 140 Mflops now and 600 Mflops in a few years. But that is not a supercomputer'. Worried, that the matter may never be cleared up, Seshagiri set forth ... on a what a supercomputer actually is. 'A supercomputer runs very large I/O capacities, vector processing capabilities, large memory storage and large terminal connectivity. Scalar processing is, equally, if not more important. Even with 140 Mflops with microprocessor (in the parallel processor computer) the speed will still only be 3 MIPS. Whereas, with a supercomputer like Cray, which can get 400 Mflops — the speed will be 60 to 100 MIPS, or more. None of us are producing this one'. ... The parallel processing project, code-named Multiple-instruction, MultipleData Stream (MIND), is part of the DOE's Fifth Generation programme, which includes such other ambitious projects as speech synthesis and speech recogni­ tion. 'Anyone who is doing parallel processing is doing it with microprocessors,' he said. 'The figure of Rs. 36 crores I have seen quoted in the press is for the entire Fifth Generation programme'. The parallel processing project is just a part of the overall strategy laid out by the DOE and it will get just a fraction of the budget. 'The budget is now Rs. 4 crores.' After Seshagiri lost his position in the DOE, K.P.P. Nambiar, the Secretary, DOE could not resist the temptation of continuing to maintain the supercomputer label for the C-DAC project without being clear about the objectives. A National Conference on Supercomputers and Applica­ tions was held in Poona, the home of C-DAC's activity, and Nambiar delivered the Presidential Address. He clarified that the C-DAC mission was to be time-bound to 3 years, on the C-DOT pattern, but with regard to the application areas, he stated:22 It could be weather forecasting or aerodynamics or remote sensing or seismic data processing or VLSI design, etc.... I am sure this conference will help in identify­ ing the priority application areas. Vijay Bhatkar, (a protege of Nambiar) was appointed to head C-DAC. The project received a massive funding of nearly Rs. 30 crores. The Governing Council of C-DAC is headed by the Minister of State for Science and Technology. C-DAC issued its first advertisement for 'Functional Directors' in June 1988. It stated: The first mission goal is to develop and deliver a supercomputer with peak

128

India and the Computer

computing power of 1000 Mflops in three years based on parallel processing architecture. Intermediate deliverables include VLSI chip sets for PCs and in­ dustry standard workstations, parallel processing workbenches for scientific/en­ gineering and AI applications, and parallel processing software products for export. The count down has begun.

One would have thought that the goals of C-DAC should have been to: develop and to bring to production world class high speed computers; to develop and bring into operation, the use of one or two major applications that significantly impact the national economy. In 1990, C-DAC put out advertisements claiming to have developed a powerful, parallel processing machine, Param and a VLSI design for a chip for the Gist (Graphics-based Information System Technology) programme. I must admit to having been confused by the advertisement blitz. R&D organizations do not as a rule advertise their successes in such a brash manner. Hawking their technology to the private sector for their exploitation also seems to underline a failure to identify any major application of national significance. Param appears to represent an entry level parallel processor built with imported boards and application software. The transputer manufacturer offers advice and assistance for using the chip, and the chip, because of its versatility, requires very little further design to make multi-node systems. Paralleling many such boards would result in larger machines. The Computer System Architects (CSA), USA, have taken to popularizing the SGS-Thomson Inmos transputer and offer boards at attractive prices. There arc similar firms in the UK as well. The Param may have been configured with such boards, along with applications software commercially available abroad acquired for demonstration purposes. The transputer project, Param, should have then been considered a start-up project, fit for helping academic institutions and colleges to learn about transputers. Its development, or assembling, cannot represent an achievement of national significance. To have spent about Rs. 20 crorcs of the allotted Rs. 30 crorcs, and to have achieved so little, speaks volumes about the utter disregard for economy: excessive hardware purchases and lavish expenditure appear to have exhausted the budget even before anything significant has been achieved. Several groups have built similar systems in India with far less expenditure of money. Guy Harriman in the UK appears to have built a far more powerful

transputer-based parallel processing computer (256 nodes)

with the assistance of a few technicians. The machine is now deployed at one of the universities in England. 2 3 The parallel processing concept has enthused several other institutions,

Mainframes — A Late Awakening who have taken up take up development of such computers. N AL was the first to announce its own design. C-DOT has taken up a project supported by the DST, and there is a DRDO project, where ECIL has a hand. 1 he Indian Institute of Science, Bangalore, is also trying to get some entrepreneurs interested in its designs. Whether all these projects will ever coalesce into a national alternative for a supercomputer has to be answered by the DOE and the Minister of State for Science and Technology. The question has to be asked, however, whether it is possible with our meagre resources, a lack of a dedicated application-oriented effort, and practically no parallel chip development, to build our own supercomputer. The software effort to deploy and use a parallel processing computer is quite significant. The tussle between Poona where hardware is being put together, and Bangalore, where the software is being developed, has led to the exit of A. Paulraj who hitherto has been directing the activities at Bangalore. Internal criticism of the present sad state of the C-DAC project may have led to his exit. A genuine supercomputer effort led by a dedicated team leader, is essential in order to make progress. China's example could serve to inspire us in India. The GIST project may produce a valuable chip for multilingual wordprocessing and use of computers in India. The effort was initiated by Seshagiri when at the helm of affairs at the DOE, and the work was carried out at IIT Kanpur.25 Development spanned the period 1982-8, in which time a PC compatible board was developed, the key person in this useful project being, Mohan Tambc. He later moved over to C-DAC to help develop some Application Specific Integrated Circuits (ASICs) to reduce board complexity. He appears to have succeeded in this, but without wafer fabrication facilities the widespread use of the GIST chip may be difficult. Chip imports can face the same hurdles of export authorization by the US Government as supercomputers do. The Space Commission is having difficulties in importing radiation-hardened devices. The 'transputer' chip developed by Inmos (now owned by the French firm Thompson) has been selected by C-DAC. Whether the transputer chip will survive the race for new generation chips remains to be seen but the transputer chip technology should at least be secured to help insure the success of the Indian parallel processing project.

130

India and the Computer ™

c. AWANtlNINU

Luhe d^3 C°m™"ee' Naras'm^an, and Menon had praclically ruled develoPment "f mainframes and large computers th i, ,n Tabounhein77 ^mmi"ce as car|y as "73 had expressed concern' throughth self-reliant in this field. The foreign exchange drain renmfy ?™ """"frames, and the difficulties in obtaining the q 1 e c earanccs from the US Government in the 1980s, forced the OE to reconsider the manufacture of mainframes with foreign technical " c o l l

n°,W"h°WdThCrC WCfC SCr'°US d C '^S

®™"zingfiw

ha^ been chosen 7 "" eVCn'' ^ Wr°"g mainframe aPPea's«° have been chosen, a supplementary agreement being necessary to obtain later vers,on. Though the DOE had repeatedly said that the markeUn lateral r SUPP°rt m°rC than °nc manufacturer, ICIM and Bull were later allowed to set up facilities, ruling out any possibility of the stand DrCTalr 0tfhl,ardWafrC 0r Sof,ware-A CKD/SKD assembly approach now prevails ,n the matnframe area. All the operating systems for the mainames, and most of application packages arc imported. The user has to pay more because of the intermediaries — the local assemblers. This late awakening has come at a time when desktop computers are challenging the matnframe: powerful PC's and workstations using RISC chips in a network,ng environment are in a position to provide a cheaper alternative to mainframes. DEC and HP are establishing themselves in puters

JUS',hiS SOrt

°f S,a,e" 0f-thC-art worksla"°ns and minicom-

melnde^llo'1 Pr°CCSSing pr°icc'-wilhoul sP«'fi= application goals will meander along as many an R&D project does in India. The basic ele' Ch'PS' r'qu,re fh""2

a" lhc

-oncy. talent, and support. But such

'aCking 'h°Ugh 3 Micr°"«"">"ic

menTco^c \ Develop­ ment Council has now been established. SCL has been ploughing a lonefy maj°r na'i0na' C"h" field or in"the '° 'He COmpmcr barriers "J^0"™"™3*10"5 ™d' " «>nti""cs to fiaun, the 'micron' arners the company has overcome, but it has been unable to contribute

SCL bchre eff07 fnlifiablc na,ional ob-Mv M c > o y 3 u a- co 4> CO c*. •~ o k— E CO O Cl U o y E ^ •M o .2 a.

3 er o 3 w fll w ^ . CL « o — E -a o s • °spS "O * ^ CL ^ O 'g -C "> CTT3 o « C *- > O Q." f, W -7 y •J C CO

« ea CO 3 e 60 « U .2 J. > ES u O 60 O aa. ea CL « ea y c O ca CL .§•? X E t> a !S S 60 *-* u. 2 C « CO CO "> JO E1 o o « CL I—• o Cl. c CL r>. i • o _ r•= E c *-* CT-TD O « C g's-3 ~ o CL ?. U Jr y « o J C CO -

«*r ,

-

CO "TO £ IT! "« «» >-• N •- U O

fs,

' .. C

O O

S L I I HCO H R -3 O 2T >

° 73 2 6 c " H 6

e

V

3 CL

E

« S?

•9 * « X CM VO 3.CN FJ CL H «RT „ (A T3 TR W 9 O

T3 JO >

F * S CM O co «-•

*

CA 4> U >•

l/l

l/l

4> >"

>

« C

e-f g" -| ~ * -g 3 CL O 3 Q..2 U CL TO "3 X) O "O >

J= *

IA , 2 £ O .2 1/1- C 5 O O •— A G N ~ FRO. G -S -a X 2 Q U 3 CL O 3 CL «• U H O CL TO 3 X O 2 >

1/5

^

IA (A IFL U U U >•>•>-

VO ^ ^ CM C) R< M N CL S

VO _

2 o "JB ^

.1 IA

* VO ^ I CM (J 00 T-. CO ££

A

IA

IA

U U U >->->-

CA T3 ° LI *

V

vo E '

I- .= TO I *R « cj — — ««

O -2 .3. I « X> * "» ^

£ A £ « ^.2 H -

i—> 00 f— ft")

X TO

- >-

O >-

(A

CL c

VO ^

» «« CM CM _ c »» i *7 2 A « W ° 8-2 £ g /> •X = p£• 2 '« .— -X « 2 Q « 5 S ; S A S - 5 H O A > £ o 2 >

IA IA IA 4» U T> >->->-

CM

l/l ~o

£> >I

O

V

? O

s

3I R-» •«* CM CO T3 l/l

2 •= «

J C

52 M

3 S B A G I Q 5 § • E 2 1" 2 2

p.. 3U30.sU H - O C L > X O O >

CM E 1— .= C S Q IUI L U H DC >

•O U °2 2 o « > 4> (A K. jS s* _. U A ^ CM .22 TCM CO -O T3

W5

« >•

T3

•n

»-

« O « > .X 9 « £ U •9 * !2 * O A * A S .2

« O CM Cl -a

V >*.§

>•>•>•

L. X.

« E

121 -C Q. U Q. rv CQU3U.tt.UJO

5 5
->->->"

§

s

fo

v, to>

|

2-

Hj co a:

u

O

o

u u

W u

(A 4>

tA 4J

CA 4>

>>>•

v£>

m m i-i 2 S S CD CD CO

oo

(A 4>

(A 4>

2 J cd

y

(A U

>• > >

a.
-

*f

s

i

Q

Q

l

1U

CM ra*

o o tj- t (A IA 4J 4) >- >•

2 2 CD CD o r-

\©

00

o-

•- CM

o

o. 2: w Q Q Q f t . D w

2 £ — O JH 3 U. a (A oo C O W ta .2 ™ i— — w a.

—

QC x: J

00 £

4) O E ° o ~ 2 — o -° 3 £ O

< 71 U

4>

o c

«-2 £ E «

E

E •- 2 £ O" 3 ^

E P

182

India and the Computer

connectors used in the computers, some of which are now being imported and which are of special design and shape. Efforts have been made progressively to achieve production indenization in this area, with coming into existence of promising entrepreneurs. ECIL foresees no major problem for indigenization as far as connectors aie concerned. Another basic requirement of computer manufacturing, namely, ICs is also gradually being indigenized with the eyelopment at BEL of some of the commonly used types of ICs. In addition LCIL and some other institutions in the country are also conducting research work Sfw Prn6"1 °fZ7 ,CS a"d Pr°Prietary ,Cs- ""»«« " good co-ordination between ECIL and BEL in the development and application of computer ICs LUL performing the required reliability tests. It may, therefore, be said that IC indenization is also well under way.... The large cost of any computer system is constituted by the peripherals which are used for input and output of information to and from the computer. As far as peripherals are concerned the country still has to make considerable leeway in indigenization. Development work ,s being carried out at ECIL, BEL, and HTL. The major challenge here is e great mechanical engineering complexity of the peripherals and the special alloys or other materials used in the manufacture of the peripherals. The cost of peripherals is high because of the sophisticated electro-mechanical devices involved as also because of expensive tooling required to make the various parts. erefore, unless offtake is high, it is not an attractive proposition for any industry to take up manufacture of peripherals. This activity will pick up only when enough demand is generated for large volume production of computers and hence peripherals. No reasonably short deadline can be given for completing indigenization of peripherals, but it is very likely that within a decade, the import fracdon

,nd,genous comPu*r

will have been brought down to a negligible

Though there were some development efforts at BEL in regard to ICs it could not be represented to the PAC as a major national effort at that' point of time. Neither could it have been the basis for the assertion that India was on the threshold of being self-reliant. It appears that the urge to project a self-reliant image, perhaps, led ECIL to be carried away. With regard to the manufacturing activity of ECIL in the area of computers, Mcnon, the Secretary, Department of Electronics informed the Public Accounts Committee: What ECIL does today is to import the peripherals, many of the ICs and memory cores. What they are doing is systems engineering- putting together the various items With many of their own items they are now making printed boards or multi-layered boards, which they did not make a few years ago. They have gone over from transistors to IC computers, they are developing software application packages but they still depend significantly on imports of those items which finally make the equipment.

Indigenous Development -The ECIL Experiment

183

On being asked whether the manufacturing operations of ECIL only amounted to the assembly of imported parts, the Secretary, Department of Electronics, stated: 1, goes beyond simple assembly for the following reasons. They are indigenizing appreciably ... but much more than that, this whole thing has got a certain design philosophy, it has got a language, it has got to have software packages. . ECIL ?s developing all these but what we want to ensure is that, even these imports such as core memory and the digital semiconductor components which go tnto >U.ke ICS and memories, must be made in India. That is what we are really aiming for, apart from the peripherals which I have mentioned. With regard to the price structure of computer^marketed by ECIL, M.G.K. Menon informed the Estimates Committee: What I would like to say is that we ate fully aware of the need to examine the costing and make sure that it is reasonable from the potnt of view of nattona usage, and this is particularly important, because we do provide a stgntficam amount of money in the form of both loan and grant to ECIL for the who e development programme, and since we provide this money, I thtnk we have also a responsibility to ensure that what comes out of it is meaningful in the nat onal context including the price structure. We will be going into this. At present h Ses are rather high This really is the consequence of the fact that all the peripherals are imported and the heavy duty which adds to the price structure. An econometric study carried out by IPAG came to the following conclusion:? ... that a computer (4K CPU) priced a. 10,000 dollars in USA should not cost more than Rs. 2.9 lakhs in India. This is the worst case estimate. However, the most likely estimate will be between Rs. 1.7 lakhs and Rs. 2 lakhs. P™^ed price of TDC-16 (4K CPU equivalent) is about Rs. 8 lakhs. It is concluded that there is a clear case for probing deeper into the costing proposals of manufacture of minicomputers in the country. ADVERSARIAL ATTITUDES As the scientists in the Electronics Commission and ECIL had the same roots it is hard to understand why a frank discussion did not take place to asccr'tain the reasons for the higher cost of ECIL systems. What were EClL's real problems? Instead of allowing an adversarial attitude to arise between a public sector body and a government department, should not the reasons for the higher costs have been explored through mutual discussion and an understanding of the real causes be arrived at- The scientists and other staff in the Electronics Comm.ss.on/DOE had no

184

India and the Computer

exposure to the realities of an industry and were unable to appreciate the problems faced by the ECIL development group, committed as it was to the development of an industry through self-reliance. In 1973-4, the question of whether all the public sector undertakings in the field of electronics should come under one umbrella — that of the Department of Electronics — was very much in the air. The pros and cons of the issue were discussed with the Estimates Committee (1973-4) where it was suggested that the Department of Electronics, as the nodal ministry should have an adequate say in and control over the vital aspects of the operations of these undertakings, that is:8 (i)

Framing the objectives of the undertakings.

(ii)

Determination of the product-mix, including diversification.

(iii) Complexion of the Board of Directors. (iv) ApjxDintmcnts to the Board of Directors and to senior management positions like Managing Directors and General Managers. (v)

Personnel and remuneration structure/policy.

(vi) Policy and follow-up relating to development of ancillaries. (vii) Pricing policies. The Estimates Committee concluded:9 The Committee recommends that the government should review at the highest level the question of placing units which are charged with the responsibility for manufacturing electronics under the Department of Electronics in the overall interest of development and for these units being used as centres of growth for an accelerated programme of development of electronics industry in the Fifth Plan.

The government did not act upon the recommendation. One wonders whether, had ECIL or BEL been placed under the DOE, if the eventual confrontation could have been avoided. Or, perhaps, the Electronics Commission/DOE may have even acquired some industrial feel which they otherwise so sadly lacked.

ECIL's PROGRESS UP TO 1978 ECIL obtained an industrial licence for the production of 50 computer systems per year. The various computers installed up to 1978 are shown in Table 8.3. Except for four computers, the other 94 were sold to government departments, universities, the police, and other government bodies. The Atomic Energy Establishments, TIFR and ECIL, alone accounted for 31 systems. This inability of ECIL to increase its production and to enter the

Indigenous Development - The ECIL Experiment

1i

Table 8.3 ECIL Production up to 1978 Year

TDC-12

1971 1972 1973 1974 1975 1976 1977 1978

1 4 5 5 5 1

Tc>tid

21

TDC-312

TDC-316

Micro-78

Total 1

4 5 7

2 9 12 9

16 21 22

10

41 3

10

98

3 32

35

SOURCE: Om Vikas and L. Ravichandran, 'Computerization in India: A Statis­ tical Review', 1PAG Journal, vol.16, 3, New Delhi, December 1978, pp. 323-32.

EDP market may have been caused by the lack of requisite software and high costs. The performance of ECIL was reviewed by the Committee on Public Undertakings, the following paragraphs being of interest: The Computer Group of ECIL incurred losses of Rs. 69.24 lakhs, Rs. 165.03 lakhs and Rs. 123.59 lakhs during 1977-8, 1978-9, and 1979-SO respectively. Asked about steps taken to improve the performance of this Group, the Corporation informed the Committee in a note that a Management Board had been constituted (1979) by the Board of Directors, to give a direction to the operations of the Computer Group and to dovetail its long range and short term programme within the framework of the existing policy. The Management Board was stated to have identified the following reasons for the poor performance of the Group: (i) With liberalized import policy, nearly Rs. 60 crores worth of computers were imported in 1977 and 1978. (ii) ECIL continued to depend on indigenous development of computer technology while technology developed by leaps and bounds abroad. (iii) New entrants in computer manufacture commenced at the lower end using imported LSI technology. (iv) ECIL had undertaken several highly complex projects, and (v) Development of their software was time consuming.

Regarding the import of computers, the Managing Director of the Corporation claimed before the committee that they could have^sold Rs. 10 crores worth of computers had imports not been allowed. While

India and the Computer

^ZZl ^

econoJaut'wta'^^.

performance:'0

f0"0wlng mcasu"* were

to ta marketed'to

ba^'/n^^ fie^pr°™

316 and 332 systems should be enhanced with th f

suggested to improve

shou,d

''

«ntin«

caPabili,y

°f the

systems of similar types, if necessary by nrn " aValIabie on 'he latest technology to cut down the development Pr°™™g0ne'mP°rte HindUS,a" Umited, etc., were to cater to essentially commercial applications Thdrke,,n8'compufer^sterns adopting the kit assembly methods bv imnnrtin completely assembled with components 'ng

fu,,y

COmpan,es were P^aps ^St^d pr,n,ed eircuit boards

import of the process. It was difficult for an • ? ™ °ds ,mplied aImost total lgenous,y manufactured product to compete with these almost im ^ OPCra,i°"S °f were stated systems being supported by these companies we"5'""53 b'S the small microcomputer to the Z j

ECIL'STDC-312, TOC-316 claw of machines

°"

PCrf0™nCe

the

°f

HC1L informed the committee in a note that IC.M, a subsidiary oflCL,

Indigenous Development - The ECU Experiment

187

had been given a licence to manufacture the 2904 series computer sys­ tems which were comparable to the medium and large scale computer systems manufactured by ECIL. In this regard the Secretary, Department of Electronics, stated that in terms of specifications, performance and volume of production, ECIL was producing a relatively small number of TDC-312 and TDC-316 computers at the time when ICL was given the licence in 1978 and that they in no way clashed with the ICL 2904 system. The committee were also informed by a representative of the department that at that time the trend in the import of computers was running at a rate of five or six systems a year and that this trend was expected to go up. So the plan was to economize on imports. ECIL's TDC-332 system was still only on paper, in that ECIL were in the process of designing it. There was also a distinct gap between the TDC-316 and the TDC-332. The Depart­ ment of Electronics wanted to fill this gap by giving a licence to an Indian company. At that time only ICIM could provide the necessary technology. The Secretary, Department of Atomic Energy, informed the com­ mittee that they objected to ICL being given a licence for manufacture of 2904 computers. A representative of the Department of Electronics ad­ mitted that ECIL had requested the government to wait for a year or so. When asked to state the reasons for not accepting ECIL's request, the witness stated that the import trend was on the increase and the Depart­ ment of Electronics assessment was that ECIL's TDC-332 programme might not be ready by March 1979. Therefore, it was thought expedient to get another production programme off the ground. In this connection, the Secretary, Department of Electronics stated: 'We are now given to under­ stand that TDC-332 can be made fully operational by the end of 1981. There would be no conflict, since TDC-332 is very powerful.' On being asked whether all these points had been discussed with the Electronics Corporation of India, a representative of the Department of Electronics informed the Committee: This particular question was examined on technical grounds and a committee was set up by the then Chairman, M.G.K. Menon. This particular committee, consisting of technical people, people from industries, academic institutions and so on, deliberated on the specifications for 2903, 2904 series with respect to ECIL's product ranges like TDC-316 and TDC-332. We knew the specifications of the systems that they already had. From that point of view, we did have discussions but they were not part of the Committee.

In reply to the question of whether the matter was discussed with the Department of Atomic Energy and whether the department was con-

188

India and the Computer

vinced, a witness stated:4 It was not discussed with the Chairman, Atomic Energy Commission, but it was discussed with the management of ECIL. They did not agree, but we made an assessment.' According to a report appearing in the journal Computer Weekly International dated 16 March 1978, ICL would in fact be manufacturing 2903s, calling them the 2904 series. When the Committee desired to know whether the Department of Electronics had taken any precautions in this regard, the Secretary of the Department stated: Whether it is a small licensed manufacturer with imported contents or a large computer manufacturer, the licensed party has to provide a phased manufacturing programme and a year by year indigenization programme. We know that they are going according to the phased manufacturing programme. If they are making any deviation, we will know it. We have a standing committee which monitors the entire thing. They cannot get away like that because they have to supply to the various people and the computer professional community in this country is very well integrated. We have an effective monitoring mechanism by which we feel we would be able to look into this aspect and at present, I have no apprehensions.

The Committee were told by ECIL that in an earlier case when ICL were given a licence to manufacture the 1901A computers, the company did not meet the obligations stipulated in the licence. The mere assembly operations resorted to by them neither resulted in any technology transfer nor in any major employment creation. The representative of the Depart­ ment of Electronics conceded this and stated: 'The facts are like that. They did not carry through the programme. They stopped in between.' When the committee tried to find out what precautions had been taken before granting ICIM the licence for the 2904 series in order to ensure that ICL would now fulfil the terms of the licence, the witness stated: Before it was approved, we had detailed discussions with the Indian company as well as the foreign collaborators, what technology they were going to have ... Government had the means to control the company and make them behave. We have used all the means in terms of approving the phased programme and then only the licence is issued.

This worry about licensing ICL/ICIM or any other private sector company to enter the market with systems basically built with imported subassemblies arose because ECIL's own systems were being priced out of the market. The private companies were able to offer cheaper solutions for many EDP applications. The DOE, faced with persistent user pres­ sure, was contemplating a policy shift — the 1978 policy announced after years of delay had not really improved the situation regarding the

Indigenous Development - The ECIL Experiment

189

availability of computer systems. The presence of very attractive systems in the world market was therefore forcing the government to look for alternative policies in 1979-80. The TDC-332, meanwhile, was still under development and there were doubts whether it would fill the requirements of a powerful mainframe which was then needed for various tasks in India. S. Manikutty, has noted:14 Both the Electronics Commission and the ECIL expected in 1971 that the 12 and 16-bit computers ... would be introduced in the market by 1974 with the full range of software needed for both scientific and EDP applications. But the technological difficulties in development proved to be enormous. The machines had to be designed from scratch ... and the problem of patents (sic) made them incompatible with the existing machines ... the software had to be developed at ECIL and by different agencies like TIFR ... which lead to considerable delays. ... Even though ECIL could put into the market TDC-312 andTDC-316by 1974, they were in fact machines with very limited capabilities. They could not be used in most EDP applications at all, because the EDP applications, by and large, needed much more software and input/output devices. ... The business language compiler COBOL was not available (ECIL developed its own version of COBOL called E-COBOL, but this being a non-standard language did not gain much popularity among users); the ECIL machines had few applications packages. ... Thus the machines introduced by ECIL were unsuitable for practically all EDP applications, and even for scientific applications their capabilities were limited. The problems faced by ECIL with regard to the 32-bit computers were still more difficult. As early as 1971, the Electronics Commission, realizing the need for large systems both in scientific and EDP applications, started applying pressure on ECIL to commence their development. The Commission hoped that by 1974, these systems would be available in the market. It was prepared to offer fund support to the extent of Rs. 2 crores for this project. But 32-bit systems were far more complex systems to develop than 16-bit systems, and the difficulties encountered in 12 and 16-bit systems led ECIL to realize that it would be necessary to master the design of a 16-bit system first before venturing into 32-bit system. Thus only in 1973 did ECIL commence the design of a 32-bit system, named System-332. ECIL had by this time also realized that if it followed the route of developing everything de novo, it would be undertaking a truly enormous task. Hence it decided to compromise. It would import a French machine, IRIS-55, that could serve as a model for hardware design, and, more important, the entire software that was available in IRIS could be straight-away used in System-332. It was thought that, by this method, ECIL could skip most of the software development; hardware development also would be much easier. This option, however, had its price. IRIS wasa system not compatible with IBM or with ECIL's own other systems, with the result that no cross utilization of

190

India and the Computer

packages was possible. The IRIS system being unfamiliar, it took a much longer time to debug the errors, and sort out the problems; and considerable work had to be done in adapting the IRIS system based on programmes and output messages in French to English. To assist ECIL in its problems, no references could be made to CII-Honeywell-Bull of France (who were the manufacturers of IRIS) for there was no formal collaboration agreement. Perhaps C-H-B was willing to look the other way, while ECIL was doing itsSystem-332 development, but it could hardly be expected to assist ECIL directly. Therefore, ECIL had to face the difficulties all alone, and these difficulties were formidable. As one ECIL executive stated, 'We just did not realize what we had taken up'. To just give the reader an idea of the magnitude of work involved in developing a 32-bit computer, it had cost US $ 5 billion, and taken five years for IBM to develop its celebrated System 360. To be sure, by importing IRIS-55, the task of ECIL was comparatively much simpler, but still was difficult enough. As a result of these difficulties, ECIL's product range in the field of mainframe computers was very limited. It had only two models in its range: TDC-312, and TDC-316 (the obsolete, second generation TDC-12 had been discontinued from 1974). Even these models had very limited capabilities in terms of equipment and software. In the field of microcomputers, major developments were taking place abroad. ECIL's response to these developments was very hesitant. Srikantan, the Head of Computer Division, and many of his colleagues did not feel that it was worth while for ECIL to take up manufacture of microcomputers. Microcomputers hardly posed the kind of challenge posed by TDC-312 or TDC-316; the value added in their production was small, and it was felt that these microcomputers would have only very limited capabilities and would never find many applications in the EDP market or the scientific market, but would be confined to stand-alone real-time applications. Only by 1975 was it appreciated that with the availability of more powerful chips, they could find possibility in the EDP field

broader applications. Even so, the

was felt to be very limited, and the Division started

developing a microcomputer, Micro-78, for scientific/real-time applications. This could hardly be completed by 1976, so that as at the end of 1976, ECIL's product range did not include a microcomputer. It could be asked whether the Computer Division would have developed its products more quickly, if there was competition. Evidently, no clear-cut answer is possible, and the ECIL executives seemed divided on the issue when questioned. Many agreed that with better co-ordination with personnel in different groups, or with product-centred teams rather than activity-centred teams as was the case, quicker development would have resulted. Pressure from competition would probably have forced ECIL to go in for products of newer technology; one executive who was in the design wing of the computer group and who has subsequently left the company stated that there was in fact not much thinking on the usage of alternative components and subassemblies that were becoming

Indigenous Development - The ECIL Experiment

191

available at lower prices. In interviews, Srikantan, the Head of the Computer Division till 1981, felt that even though looking back one could say that so many things could have been done, at that time they did what they thought was the best. A.S. Rao himself felt that though the Division did very well till 1974, a slackness set in then, and much more could, and should have been achieved. In his view, the Division could have easily achieved and even surpassed its target if it had exerted more. It cannot, however, be disputed that there was a significant shortfall as regards product development, and this contributed to the inability of the Division to attain its target in sales.

A.S. RAO RETIRES A.S. Rao, the moving spirit behind ECIL and a devoted adherent to the philosophy of self-reliance, retired in 1978, somewhat dispirited because of setbacks and also because his views no longer commanded attention at Delhi. Disillusionment had, perhaps, set in even earlier — in 1974 when ECIL's performance was not up to his own expectations. He was reported to have asked all his senior colleagues to go on leave in March 1974, so that he could reflect on the position. After a week, he called them all back and even promoted some; but to many, the working atmosphere within ECIL was never the same again and the easy dialogue with A.S. Rao disappeared. Self-reliance also requires a good measure of self-denial. Om Vikas noted:15 Self-reliance requires not only determination but also the strength to develop and use the capacity for autonomous decision making on all aspects of development ... technological autonomy does not mean isolation although it may involve selective de-linking from the world market systems for periods of time or sector of economic activity while indigenous problem solving capabilities are being tested and developed. The developed countries during their industrialization, had erected high tariff walls to protect their nascent industries — obviously a clear case of selective de-linking.

But the policy makers at Delhi were no longer able to carry the political leadership or withstand the lobby pressure from various other groups complaining about the slow progress in the field of computers. Sarabhai had selected S.R. Vijayakar (Commercial Superintendent of the Ahmedabad Electricity Company, a mechanical and electrical en­ gineer) as the General Manager of ECIL, a post Vijayakar occupied till A.S. Rao retired. But as General Manager, Vijayakar, was confined to administrative duties. All the technical managers reported directly to

192

India and the Computer

A.S. Rao. The General Manager was mainly responsible for personnel, purchase, and public relations. The technical managers took most of the decisions themselves or consulted A.S. Rao. The General Manager was nowhere in the picture. When Rao retired, Vijayakar became the Manag­ ing Director. Vijayakar was not wedded to the idea of self-reliance nor did he have the missionary zeal of A.S. Rao. As a manager, he wanted to improve the balance sheet and adopted the tactics of the private sector in doing so. He decided to import technology and to undertake phased manufacturing programmes — CKD/SKD to start with. As he approached the DOE as a supplicant, he was welcome. On the other hand, the injured stance of A.S. Rao pointing out the fall of standards in relation to the concepts of self-reliance — the basic concepts which had brought into being the Electronics Commission — had made him unwelcome, especially as powerful trade interests had subverted various officials of the DOE who were more receptive to quick-fix ideas rather than towards genuine industrial development. S. Manikutty in his dissertation on ECIL states:16 A major function of general management has been noted in the business policy literature as that of ensuring 'a match' on a continuing basis between characteristics of the organization and those of the environment.... In the early years, the government gave ECIL protection from imports and competition. In 1976-8, the government liberalized its import policies for computers and licensed competitors to manufacture computers. ECIL response was mainly to attempt to influence the policies to nullify the effect of these changes. Its ability to do so, however, turned out to be limited, because ECIL had not been able to come up with a satisfactory range of computers. ... Nor was there any significant changes in overall strategies of the Corporation. This lack of response could be traced to a strong commitment to certain values on the part of its Chief Executive. S. Manikutty, states in his conclusions, among other things: Public Enterprises whose objectives are based on values and whose top management is deeply committed to these values may have difficulties in appreciating the need for strategic changes if it implies change in these values. Thus commitment to values can be a barrier to strategic changes. He further adds: 'Change in leadership can bring about a different appreciation of the environment and hence a strategic change.' In public service, commitment to values have been considered important. Mrs Indira Gandhi believed and encouraged 'committed' public servants. If values change at a political level, it is indeed difficult for a committed

Indigenous Development - The ECIL Experiment

193

public servant to change his convictions. The failure of ECIL to achieve acceptable results in the computer field is the failure of the basic premise of self-reliant development — in a limited time-frame with limited resources — in a rapidly changing tech­ nological field where enormous sums of money and effort were being expended elsewhere in the world. There has been no public acknow­ ledgement of this either by the scientific community or the political leadership. The public posture has always been one of commitment to self-reliant development. Denigration of the public sector for pursuing a role publicly accepted by the political leadership but privately doubted, led to the disenchantment of many of the engineers working in ECIL. Many left to join private sector companies which offered better prospects. Indeed, Srikantan left in 1981 to take up the Chairmanship of the Andhra Pradesh Electronics Development Corporation. Vijayakar did little to improve the sagging morale. At least A.S. Rao had remained intellectual­ ly honest to the very end, clinging to the values which were being publicly proclaimed and acclaimed but privately doubted and denigrated. But to judge the performance of ECIL, or for that matter that of any other public enterprise, strictly on business standards, is not valid. In spite of its many difficulties, ECIL managed to deliver quite a few computer systems for scientific applications and for use by various government departments (sec Tables 8.4 and 8.5). 1 7 Note that very few systems were sold to private parties. ECIL's main market lay with government departments, co-operatives, public sector, universities, etc. A summary of the investments made by ECIL in the computer area year by year, and the profit/loss position of the computer division are shown in Table 8.6. 1 8 Table 8.4 Total ECIL Computer Sales to Private Parties Private Sector Sales

TDC-12 TDC-312 TDC-316 Micro-78 Micro-32 SOURCE:

1

8 9

8

Data from ECIL.

Total Systems Produced

19 48 141 128 138

194

India and the Computer Table 8.5 ECIL Computer Deliveries up to (1986-7)

Type

Qty.

Amount (Rs. lakhs)

Year

Delivery Qty.

Value

TDC-12

19

218.63

1970-1 1971-2 1972-3 1982-3

1 5 12 1

8.97 66.46 136.09 7.11

TDC-312

48

632.93

1973-4 1974-5 1975-6 1976-7 1977-8 1978-9 1979-80 1980-1 1981-2

8 6 6 9 2 2 8 6 1

64.59 73.33 43.03 152.81 36.66 42.76 134.65 77.46 7.64

TDC-316

141

4,192.91

1973-4 1974-5 1975-6 1976-7 1977-8 1978-9 1979-80 1980-1 1981-2 1982-3 1983-4 1984-5 1985-6 1986-7

2 7 9 10 6 3 6 5 12 23 15 12 28 3

33.65 197.49 266.39 269.17 209.55 45.80 147.17 117.63 345.38 496.90 561.09 299.32 711.12 492.25

System-332

20

2 4 6 3 5

91.42 150.85 234.36 87.76 157.05

Total

228

5,765.91

Micro 78

128

444.29

1976-7 1977-8 1978-9 1979-80 1980-1 1981-2 1982-3 1983-4 1984-5

8 13 25 13 12 26 6 13 12

13.90 44.80 64.93 25.78 37.38 151.80 26.35 46.32 33.03

Micro 32

138

986.02

1983-4 1984-5 1985-6 1986-7

1 17 61 59

11.70 138.98 409.47 425.87

Total

266

1,430.31

721.44

1981-2 1982-3 1983-4 1984-5 1986-7

Indigenous Development -The ECIL Experiment

195

Table 8.5 (continued) ECIL Computer Deliveries up to (1986-7) System Services

Qty.

MTDES

Delivery Qty.

Amount (Rs. lakhs)

Year

11

160.62

1981-2 1982-3 1983-4

5 3 3

53.75 48.95 57.92

SFT(M-32)

25

258.68

1981-2 1982-3 1983-4 1984-5 1985-6 1986-7

1 1 12 8 1 2

6.06 70.84 108.31 52.84 6.63 14.00

Work Stations

8

124.61

1985-6 1986-7

1 7

2.30 122.31

Teletext SFMSS(M-32)

1 1

119.25 62.52

1985-6 1986-7

1 1

119.25 62.52

993.74

1973-4 1974-5 1975-6 1976-7 1977-8 1978-9 1979-80 1980-1 1981-2 1982-3 1983-4 1984-5 1985-6 1986-7

10.12 30.53 76.55 68.27 87.55 83.35 147.05 56.67 75.46 77.31 54.70 50.87 82.12 93.19

2,931.17

1975-6 1976-7 1977-8 1978-9 1979-80 1980-1 1981-2 1982-3 1983-4 1984-5 1985-6 1986-7

21.68 68.79 77.35 131.73 175.99 169.07 199.97 280.50 340.22 387.39 472.14 606.34

Enhancements

Services

Total

72

Value

4,650.59

NOTE: MTDES — multi-terminal data entry systems. SFT (M-32) — for com­ munication purposes. SFMSS (M-32) — for communication purposes. SOURCE: Data from ECIL.

196

India and the Computer Table 8.6 ECIL-Investments/Profit/Loss (1970-87) (Rs. lakhs)

Year 1970-1 1971-2 1972-3 1973-4 1974-5 1975-6 1976-7 1977-8 1978-9 1979-80 1980-1 1981-2 1982-3 1983-4 1984-5 1985-6 1986-7 Total

Investments

Profit/Loss

397.64 98.70 126.55 47.00 81.61 161.90 65.70 (2.04) 8.78 80.62 41.08 (75.59) 274.16

(0.49) 21.58 19.56 21.81 37.18 (29.02) (35.97) (69.24) (165.03) (123.59) (199.34) (26.61) 53.49 (56.62) (221.92) 81.40 179.21

1,306.11

(513.60)

£

* Total inclusive up to 1974-5. SOURCE: Based on a personal communication from J. V. Kanitkar, DGM, Corporate Planning, ECIL, 5 February 1988.

It will be seen that up to 1986-7, ECIL had invested a sum of Rs. 13.06 crorcs in the computer division, delivered computer systems and services amounting to Rs. 118.46 crores, and incurred a marginal loss of Rs. 5.13 crores. The computer division of ECIL had a staff strength of 1208 as on 1 February 1988, made up of 710 technical and 498 non-technical personnel. No doubt they had also received grants from the DOE and the DAE for design and development, but their delivery in financial terms had been significant, though one could argue over whether prices charged were fair. The development of software was from the beginning a major task. Developing a major computer system backed by fully supported in-house software is not a task that can be carried out in a short time. That ECIL could partially succeed within the time-frame and the resources available was indeed commendable. ECIL, assisted by TIFR, and to some extent by IIT Kanpur, put in considerable effort and did succeed in producing software (see Table 8.7) for the TDC scries. In addition, ECIL developed custom-built software packages for:

Indigenous Development - The ECIL Experiment 1

Data acquisition systems for the Atomic Energy Department.

2.

Satellite launch vehicle monitoring, telemetry, tracking for space

3.

applications. Rolling mill data loggers, billet cutting system, coiling system, etc., for the steel industry.

4.

Engine testing for gas turbines and launch vehicles.

5.

Several packages for message switching and communication sys­

6.

tems. Enquiries systems for crime, criminal investigation and directory enquiry system for the telephone department.

7.

Several packages for banking, insurance, dairy, oil, ordnance fac­

tories and coal. It was unfortunate that the software was not made upwardly compatible and that this lack of compatibility forced ECIL into avoidable duplication of efforts. The computer architecture did not take into account the neces­ sity for the compatibility of system software. Scshagiri, (the then Addi­ tional Secretary, DOE,) told Computers Today: ... but I tell you what happened to the indigenous development of the operating system by ECIL as a case study. See, when TDC-312 and 316 were being developed they said that they would not buy software, they would just develop it locally because of this great ideological principle that since software is manpower-intensive, it is our duty to develop it locally. Let us examine the penalty of this misplaced patriotism. ECIL recruited 100 software specialists and then took three years to develop the software for TDC-312, another three for TDC-316, and another three for TDC-332, by which time TDC-312 became obsolete — they could not sell as much as they could have otherwise. This happened for the other two systems also. They were actually playing a lag game. Even when they did reverse engineering (IRIS), it took three years, and at the end, the total cost of the system was such that the total foreign exchange component was more than the cost of equivalent imported system! Is this the price one pays for indigenization? So there should be some higher rationale governing the whole thing. What is wrong in buying technology? What is wrong in buying source code know-how for system software? (emphasis added) The 'they' all belonged to the TIFR fraternity of which Scshagiri was also a member. The TDC Working Group was chaired by P.V.S. Rao and with Scshagiri looking after all the TDC work. The question has to be asked whether a dialogue with ECIL and A.S. Rao could have been possible. The Electronics Commission could have discussed the matter openly. t t The fault lay with not only with the 'they' but also with the 'we .

c

o

o .E I s l a. . o - = u = E o oO .5 o oo 2 3a 1 o jCo c c3 oW 60 c2 co3 o 0 o 'o> «u >» b. ra a. -C v)^ Z« J CL c j= o a.io w

o 5 12 5 CQ Q a: «60 5?. .5 T3 •£ O.C a." ™ » 2 _) 5 z Q

= = S] L b.5 Uo >u

wou

QQQ

o | Q 2m

E COc ts« ° E •/I O (N W cW 3- "P « r* el IE

O bT E feo -oa.-SC w

X .2r- ..2 UUJ O =3

P

5b M /g

00 >[™ >> Q i/i t— i 0 Q F

y 2 5 C-2 E

7 o Q H

.2 -S 2'-=

S> 6 ; p " E u ;

200

India and the Computer

Perhaps, Seshagiri got carried away by the necessity to sell his 1986 policy initiatives to his interviewers. ECIL's role in the area of computers did not come to an end, but the goal of self-reliant development was abandoned. Faced as it was with basic changes in national outlook, and Vijayakar's desire to put ECIL back in place as a leader in computers, even if it be with imported know-how, global tenders were floated for technology for a super­ minicomputer. The DOE floated another tender for a mainframe, Vijayakar seeing to it that mainframe manufacture was not taken away from ECIL. Norsk Data was selected to supply the supermini — a com­ pany which started operations at about the same time as ECIL (1967) but which had made phenomenal progress, untrammelled as it was by redtape.

NORSK DATA — A CONTRAST It may be instructive to examine the ingredients which made Norsk Data a success. The story of Norsk Data has been documented in 0yvind Hcradstvcit's, Norsk Data: A Success Story. A brief summary will serve to highlight the growth of the company.20 Norsk Data was founded by three technocrats: Lars Monrad-Krohn, Per Bjorge, and Rolf Skar. Lars Monrad-Krohn was a graduate of the Norwegian Institute of Technology and the MIT, USA. Before starting Norsk he was an employee of the Norwegian Defence Research Establishment (NDRE), where he had worked on the third generation computer SAM2. Per Bjorge, was also an alumni of the MIT, USA, and the young Rolf Skar later went on to the same academic establishment. The dedication of the team was enormous. One of the team members, Skar, talking about the development of the SAM2 at NDRE said: 'We worked like crazy, 24-hours a day! One group started at 10 in the morning and worked until 2 next morning. Group two showed up at 20 hrs in the evening; they did not leave until they had at least one hour with group one the morning after. In other words, each group worked about 14 hours at a stretch.' The SAM2 computer which they designed was a third generation computer using integrated circuits and was probably among the first of its type anywhere in the world. The company was founded on 7 July 1967. The First computer NORD-1 was installed on the ship Taimyr in 1968. The NORD-1 computer could

Indigenous Development - The ECU Experiment execute 500,000 instructions per second. The system was designed to protect ships against collision. The computer was to process information from the radar, register traffic in the ship's vicinity, and calculate its course and speed. With this system, the bridge would have a complete picture of all nearby traffic and thus be able to avoid collisions. In addition to the anti-collision system, there were five other functions that the computer system could support; (1) Regulating power supply on board, (2) watching over the ship's machinery, (3) giving notice of any irregularities, (4) calculating the optimal distribution of cargo, (5) and keeping track of the crew's wages. In short, a total system for a ship. And Norsk Data got the order with a resource of only three highly talented computer scientists. Five men from the project group went with the ship on its trial run to the Far East. The ship returned four months later, and it was reported that the computer had functioned as planned, and Norway had become a pioneer in this field. The most significant thing for India to note is that a computer for ship automation had been designed, developed, produced and delivered within one year, including the provision of all necessary software. The total number of people employed at Norsk Data was 13 only at that time. The infrastructure in Norway was in no way superior to that in India. Norsk bought all the components from the best available source, assembled and configured the system. The whole team are reported to have worked around the clock in order to effect the first delivery. Other management experts and businessmen joined the Board of Norsk Data to provide support. In 1971, 88 people were employed by Norsk Data and the sales had increased to over 15 million kroner (about Rs. 3 crores); the capital stood at 3 million kroner; profits about half a million kroner. In February 1971, Norsk Data won the contract with the Meteorological Institute in Norway in competition with several large international companies. Their Nordic system was ready in April 1972 and was claimed to be the first 32-bit minicomputer in the world. The value of the system sold to the Meteorological Institute was about 6 million kroner. Explaining how the work for the Meteorological Institute was completed, Skar states that they moved all their material to the work place and worked day and night. He notes that the work rarely ended before 2 a.m. A core group of 25 people were engaged on the Nordic project. After strengthening the marketing organization, Norsk Data continued to progress. By 1973 Norsk Data had delivered 105 of its systems for ship automation. In December 1972, they won an order for computer systems

202

India and the Computer

for CERN (the European Centre for Nuclear Research) against interna­ tional competition. The contract with CERN led to the delivery of 24 computers worth 12 million kroner; and by 1976, Norsk Data had sup­ plied 50 computers to CERN alone and the company had become a strong European alternative for the supply of minicomputers which could com­ pete with the big American minicomputer suppliers. The other major area where Norsk Data made a notable entry was in the field of supplying computer and communications packages. The first such computer package was produced for meteorological forecasting in Algeria. The contract came through a United Nations Organization — the International Organization for Meteorology in Geneva; the contract was worth 14 million kroner. The NORD-5 system was stated to be as powerful as computers that were sold for three times the price. Norsk Data was also able to obtain the contract for the simulation computer for the F-16 (US fighter) programme. In ten years, the company had reached a sales turnover of 112 million kroner (about Rs. 22 crores). The company had grown at the rate of nearly 45 per cent per year. The total number of employees was only 279. Out of these, 55 were university graduates with science or engineering degrees, 90 had higher engineering or other university degrees, 33 came from vocational schools or commercial colleges, 25 were educated in secretarial colleges, and 18 were engaged in positions where educational qualifications were not required. The profit was 6.5 per cent of the operating revenue; exports represented almost 48 per cent of sales. Greater attention came to be paid to marketing and raising further capital and the company continued to grow at a fantastic rate, reaching a turnover of nearly 2500 million kroner (about Rs. 500 crores) in 1986. Norsk Data had then become a computer company with an international standing and prestige. The total number of employees stood at 3569 in 1986 and the share prices had skyrocketed and the company's shares were listed on London, New York and other stock exchanges. If one compares the success story of Norsk Data with the failure of ECIL (or for that matter with any other Indian high-tech company), the stark contrast is clear. The Norwegian company was started by three pioneers and grew because of state-of-art of the products and the extreme hard work put in by all the staff. Obviously, the company must have bought all the components, including memories and all the peripherals. It is not known whether they had their own PCB facilities in the early days. The greatest advantage they had was the intercourse with the USA and the European computer industry. They did not isolate themselves. There

Indigenous Development — The ECIL Experiment

203

were no import control procedures or stiff import duties; they did not have to contend with rigid rules and regulations causing delays or with foreign exchange problems. They could operate in an atmosphere conducive to free enterprise. Norsk are currently facing problems because their computers were designed on their own non-standard operating system. The need for Unix compatibility, the difficulty in using standard packages, networking problems, and the availability of cheaper desktops have all made market­ ing difficult, and the company has run into losses. It has tried to retrieve its position by a change of strategy, but in the fast moving world of information technology such changes of fortune are commonplace even for established companies. In view of rapid developments in the field of computer technology, production engineering and software, it was necessary for ECIL to have kept abreast of the developments abroad so that their progress could have been measured against world trends. Neither A.S. Rao or Srikantan found time to visit any other computer factories or laboratories. Rao mentioned that he did visit Fujitsu in Japan once along with Sarabhai in the late 1960s; his impression was that at that time Fujitsu was operating at about the same level of technical competence as ECIL. Today, Fujitsu rivals IBM in technical excellence and market domination. Though some of the junior scientists of ECIL did visit factories and laboratories abroad, the main input for the top management-was journals and magazines. Srikan­ tan could have acquainted himself with international trends and advised Rao, but he had to contend with the twin problems of guiding develop­ ment and setting up a production operation. ECIL had a laboratory culture and not an engineering one. Though BEL had acquired an excellent production culture through its collaborations, ECIL's environment remained that of a prototype shop.

ECIL's FUTURE ECIL still has to face problems in the computer field. The mainframe area has been thrown open to other entrants and the government have selected the CDC computer with an eye to scientific applications. There have already been serious delays in the implementation of the project because of change of the CDC model from the 830 to the 930/932 — though through no fault of ECIL. The assembly of 930s has now commenced at Cherlapally, Hyderabad, but ECIL may have to live with a govern­ ment/public sector market subject to budget cuts. The Norsk Supermini

204

India and the Computer

has also now become more or less obsolete with the arrival of powerful desktops and workstations with the Intel 80386/486 and RISC chips ^"mPU,er Dlvlslon inn M HK

T

(^TableOr

recorde 3

co ON JU -O

£

r~

rJ 2

MO 00 1r» po — ~ •0 tj^

MO lO « S 2 ^ S

rf

rJ

PO t _

J ^ M ^ O MO t n O •— t •— h v> n • P O

82

CJ fN (S M

.— >—

wo 5

T O 2£j

00 Of o —M

$

8 S

- 8« 2

q.5 E

!'§ "8

m

H

s §.g i g § •§ *• a = a 3 « a c o , < ' j 2 5sfl«a\SJ3li25«S

^ c.-2 E o E o uo O

•§$$ §

II IS > H

217

The CMC Ltd. -A Qualified Success

It may be noted that expenditure has kept up with the income leaving profits at almost the same level. Table 9.4 summarizes the financial position of the company under broad headings. The debt-equity gearing of this company is worth noting— 0.58:1 in 1987-8, 0.62:1 in 1988-9, 0.67:1 in 1989-90, The interest burden on the operations has been greatly reduced by the very liberal equity funding of the government and by loans at favourable rates. On 31 March 1990 the paid-up capital stood at Rs. 15 crores, reserves at Rs. 9.13 crores, and the gross fixed assets at Rs. 41.98 crores. Table 9.5 summarizes the poor profitability of the company. Table 9.4 CMC Ltd. — Financial Position (31 March 1990) (Rs. lakhs)

Liabilities

a. b. c.

d.

Paid-up capital Reserves and surplus Borrowings from Govt, of India from banks from others (FD) Trade dues and current liabilities (including provisions) Total

^Gross block f. Less depreciation g. Net fixed assets h. Capital work in progress i. Investment i. Current assets, loans and advances k. Miscellaneous expenses (not written off) Total Capital employed Net worth

1

012 56

un9S

n 1« 3,847.56 10.401'39

4,198.02 1 ° 'U q 2»51 . n nn ™ 70 39

10>401-39

S'So ?7 2,342.17

NOTE: Capital employed represents the total of net fixed assets, capital workin-progress and working capital. Net worth represents paid-up capital plus reser­ ves and surplus less intangible assets. SOURCE: CMC Ltd., Fourteenth Annual Report^ 1989-90, 1990, p. 37.

218

India and the Computer Table 9.5 Percentage Profit before Tax (1987-90) 1987-8

To sales To gross fixed assets To capital employed To net worth To equity capital SOURCE:

2.49 6.72 4.46 7.87 11.88

1988-9

1989-90

0.11 0.33 0.19 0.50 0.75

0.88 2.56 1.66 4.60 7.18

CMC Ltd., Fourteenth Annual Report, 1989-90, 1990, p. 39.

The company has only once in its history paid a dividend to its share holders — the Government of India — a sum of Rs. 25 lakhs in 1981-2. The company's expenses have always kept pace with its income and the margins have been slender. For a company in the service sector, this financial performance is poor, and worse, it is deteriorating. The Com­ pany is reported to have incurred a substantial loss in 1990-1. Either the margins have been purposely kept low, or the expenses have been unduly high (which would surely indicate a management with free-spending ways). The company has been pleading that the loans given by govern­ ment should be converted to equity so that the interest burden is lowered. But such a plea from a service company is not tenable; the company has to levy charges commensurate with its costs with a fair return to its shareholders — the government. As it is, the debt—equity gearing is favourable to the company. R&D expenditure has not been separately quantified, and the argument that R&D expenses has reduced profitability is not tenable, as development work has been project-oriented, and hence, must be charged to the projects and recovered from the customers. There were 923 systems under maintenance by CMC as of 31 March 1989. In addition, 5277 BBC Micros for the Class programme have also been maintained by CMC. Table 9.6 indicates the variety that CMC has to contend with.

CMC — IN THE SEVENTH PLAN Table 9.7 exhibits the Seventh Plan revenue projections for CMC as published in its Ninth Annual Report. As the turnover in 1989-90 was only Rs. 124 crores, it would appear that CMC had set for itself an unrealistic target in theSeventh Plan. Indonet, systems consultancy, and exports have not since grown as was then anticipated.

The CMC Ltd. -A Qualified Success

219

Table 9.6 Systems and Equipment under CMC Support (1990) Manufacturer

No. of Systems

Bull Automex (CMC) Control Data DEC 10/20 DEC VAX/MicroVAX DEC PDP 11 EC1L (332, MEDHA 830) GCS Hewlett-Packard IBM Impact Norsk Data Siemens Tata-Elxsi Unisys-Burroughs Univac Other systems

12 35 09 04 153 177 05 19 232 30 60 08 05 02 32 16 124

Total

923

'Class' computers

5,277

SOURCE: CMC Ltd., Fourteenth Annual Report, 1989-90, 1990, p. 10. Table 9.7 CMC Seventh Plan Targets - Revenue Projections (1989-90) (Rs. crores)

Field engineering operations Site preparation Indonet Systems consultancy Turnkey projects Special products Education and training Exports Grants/miscellaneous Total

76.00 8.50 60.00 20.00 28.00 20.00 5.50 30.00 2.00 250.00

SOURCE: CMC Ltd ..Ninth Annual Report, 1984-5, 1985, p. 14.

220

India and the Computer

For the Eighth Plan, CMC has safely avoided setting itself any financial targets; instead it talks of globalization of its operations and taking up 'quality-of-life' projects. A memorandum of understanding between government (the DOE) and CMC Ltd., defining the corporate objectives is probably necessary in order to clarify the future role of the company. It is essential to identify the objectives in order to ensure that CMC serves the public in the manner originally envisaged. Though CMC has achieved the initial tasks set for it, in changing circumstances, it has since set for itself other tasks of national relevance. Should it burden itself with the Class programme, for instance, as small computers can be installed and maintained by others, or by SCL itself ? A measure of fiscal discipline has to be brought about; charges for the services must be commensurate with costs, and the company must, as a matter of principle, start paying dividends to the government. Indonet should pay for itself. Though the technical success of CMC may be uncomfortable for the private sector companies, who continue to seek to restrict the role of CMC, this should not deflect CMC from its set tasks. The government, the departmental undertakings of the government and public sector should be encouraged to use CMC in their various schemes of computerization, and CMC should be conscious of its responsibilities. Gupta, in his speech at the twelfth annual meeting of CMC, has complained of the practice of 'tendering' for contracts from the public sector where it is expected to openly compete with private com­ panies. Surely an organization which was set up by the government for a specific purpose should be used by it for that purpose. NOTES 1.

Government of India, Department of Electronics, Annual Report, 1974-5, New Delhi, 1975, pp. 58-60.

2.

Government of India, Department of Electronics, Annual Report, 1973-4, New Delhi, 1974, pp. 14-5.

3.

Government of India, Department of Electronics, Annual Report, 1977-8, New Delhi, 1978, pp. 155-61.

4.

Government of India, Report of the Review Committee on Electronics, (Sondhi Committee), New Delhi, September 1979, pp. 65-6, para 12.15.

5.

Ibid., para 12.16.

6.

'Sondhi Committee and Menon Committee Reports: Government's Decision\ IPAG Journal, vol. 8, 8, New Delhi, May 1981, p. 605.

The CMC Ltd. - A Qualified Success

221

7.

Ibid.

8.

The notable projects being undertaken by CMC Ltd. include: the Nava Sheva Port computerization, the Bombay Stock exchange computerization, and the computerization of railway reservation systems at various centres.

CHAPTER X

Technology and Applications Development

DIRECTIONS The Electronics Commission and the DOE had been set up to promote technological and industrial development in the field of electronics based primarily on the self-reliance concept. Exceptions were to be made only in some areas like computer peripherals and components where technol­ ogy import was considered advisable. Menon had to set in motion steps to achieve self-reliance and promote industrial development based on these difficult concepts. The Estimates Committee (1973-4) was fur­ nished with a list of projects to be taken up in the area of computers.1 As the organizational set-up to execute and monitor the projects was still in the process of being finalized, the DOE found itself unable to spend the money allocated to it for R&D. The Estimates Committee commented on this difficulty:2 The Committee would ...like the Department of Electronics and all others concerned to make a concerted effort to see that priority schemes for research and development are at least now taken up without further delay and ensure that targets set are achieved.

The Public Accounts Committee (1975-6) under the Chairmanship of H.N. Mukcrjee also devoted considerable time to examining the in­ digenous computer industry and the associated R&D. They noted:3 For attaining the goal of self-reliance in the field of computers, the Committee feel that there is need for intensifying Research and Development. ... The Committee

Technology and Applications Development

223

recommend that government should draw up a plan outlining the general strategy self-reliance in computers, and in particular, the manner in which it is proposed to meet the anticipated large requirements of minicomputers— The Committee recommend ... that government should work out a comprehensive policy for self-reliance in the computer field, mentioning in particular how the requirements for computers in the Fifth Plan, specially of minicomputers, are proposed to be met. This, it is suggested, may be laid before Parliament in a White Paper within a period of six months of the presentation of this Report. There might then be a pointed and purposeful national debate on the subject. for

The long-delayed policy tabled in March 1978 did not specifically address itself to self-reliance and there was no 'purposeful national debate'. During his tenure as Chairman of the Electronics Commission, and Secretary, DOE, Menon decided to promote the technology and applica­ tions development of computers by setting up the following organiza­ tions: 1. The Technology Development Council with various Working Groups for different sectors. The group for computers was headed by P.V.S. Rao and was to evolve plans for technology development based on the principle of self-reliance. The Electronics Commission was to review the plans, fund development, and ensure progress. 2. Regional computer centres. 3. National Centre for Software Development and Computing Tech­ niques. 4. The National Informatics Centre. 5. The National Automation and Control Programme to be run by the National Automation and Control Agency (NACA). This was later rechristened the 'Appropriate Automation Programme'. Those who followed Menon as Chairman of the Electronics Com­ mission and Secretary, DOE, persisted with the same development schemes though they were forced to give up the concept of self-reliant development. They added a variety of other programmes, like Class (for the introduction of computers in schools) and other projects aimed at training in the area of software development (computer-aided design, computer-aided manufacture, computer-aided management, etc.) and be­ came more concerned with applications development than with hardware development. A fifth generation computer development programme, a network to connect educational institutions (ERNET), and an ambitious programme called C-DAC (see Chapter 5) which set out to develop parallel-processing

224

India and the Computer

computers which would make it unnecessary to buy supercomputers from the USA, were also launched. The UNDP provided assistance to many of these schemes. In addition, manpower programmes designed to meet the shortage of manpower in the computer area were also initiated.

TECHNOLOGY DEVELOPMENT — THE WORKING GROUP REPORT A comprehensive paper on Technology Development was prepared by a working group under P.V.S. Rao.4 The working group had nine members including Nag who later became the Chairman and Secretary, DOE. The plan covered the entire area of computers and their components, indicated both phasing and priorities, and stressed that no additional funding need be provided for the development of mini and midi processors of a general purpose type. Apparently, the working group was highly confident of the success of the ECIL programme. The working group noted:5 Standardization in the field of computers is important to achieve uniformity and compatibility and applies in particular to the peripherals area. Standardization of interfaces, communication buses, communication protocols to maintain maximum growth potential and interchangeability of subsystems is essential. ... Evaluation of standards for systems (local/imported), hardware and software, also needs to be taken up soon. Though a comprehensive paper was prepared by the working group, the financial requirements to see the plan through were not worked out. Any person aware of the international trends in the expenditure on the development of computers and their associated peripherals and software would have realized that an immense effort backed by enormous sums of money would be needed to achieve the progress envisaged by the report in so short a period of time. A total allocation of Rs. 20 crores by the Planning Commission for the Fifth Plan for the entire electronics field should have alerted the TDC and the various working groups that dif­ ferent strategies had to be evolved, if India was to progress in the field and use computers for national advancement. The Working Group did not work out an action plan either; that was left to the DOE. The report ended with the statement: It is hoped that the broad indication of areas of interest and their inter se priorities as outlined in the report would be useful in generating project proposals of a type that would be relevant to the immediate needs of India.

Technology and Applications Development

225

Table 10.1 Technology Development Completed Projects (1971 to 1983-4) Item

Software

Amount (Rs. lakhs) 485.78

Recipients (Rs. lakhs) ECIL Rs. 344.76, IlSc Rs. 5.29, TIFR Rs. 133.28, IIT Rs. 2.45 IIM Rs. 11.4, PRL Rs.14.98, Jadhavpur Univ., Rs. 1.73

Training and education

28.11

Telecom

44.27

HTL Rs. 13.46, ECIL Rs. 30.81

Hindi/devanagiri

18.96

IIT Rs. 3.69, BITS Rs. 6.77, IIT Rs. 8.50 .

729.41

TIFR Rs. 707.64, ECIL Rs. 21.77

Defence

Peripherals

91.05

ECIL Rs. 32.00, BEL Rs. 45.20, IIT Rs. 2.50, KSEDC Rs. 8.35, IIT Rs. 3.00

Hardware

43.00

Jadhavpur Univ., Rs. 3.52, ECIL Rs. 31.53, TIFR Rs. 7.95

Components

16.70

CEL Rs. 16.70

Systems

78.25

CMC Rs. 78.25

103.79

SCL Rs. 103.79

SCL Total

1,639.32 lakhs

SOURCE: Compiled from data furnished by the DOE.

Tables 10.1 and 10.2 summarize the expenditure. The rate of disburse­ ment of funds and the number of on-going projects is an index of the slow tempo of development and the lack of dedicated effort. During the period 1971-2 to 1983-4, a sum of Rs. 16.39 crores was disbursed for 31 projects completed during that time and an additional

226

India and the Computer Table 10.2 Technology Development - Ongoing Projects (Incomplete as on 31 March 1984)

Item

Amount (Rs. lakhs)

Recipients (Rs. lakhs)

Software

66.63

IIT Rs. 9.57, IIT Rs. 5.54, ERDC Rs. 16.92, SRI Rs. 10.70, Punjab Univ., Rs. 9.40, IIT Madras Rs. 14.50

Hardware

282.23

IIT Rs.16.70, TIFR Rs.16.20, INCOR Rs. 9.37, CEERI Rs. 27.63, IISc Bangalore Rs. 19.42, CPRI Bangalore Rs. 6.16, CMC Hyderabad Rs. 186.75

Training and education

51.26

JK. Institute Rs. 8.65, BITS Rs. 13.21, Andhra Rs. 9.50, IIT Kharagpur Rs. 19.90 •

Telecom

30.81

ECIL Rs. 30.81

Total SOURCE:

430.93 lakhs Compiled from data furnished by the DOE.

Rs. 4.30 crores for 18 projects which were yet to be completed. The initiative to generate project proposals was left to the various laboratories, industries, and entrepreneurs. The working group and the DOE were merely to act as examiners to vet, approve, or reject a proposal. There was to be no direct involvement. It is surprising that the TDC working group did not review the ECIL's programme of computer development or suggest any special strategies for more rapid development. The TDC progress report published by the IPAG, for example, indicated a time period of seven years for the software development for the TDC-316 and TDC-332 computers. The time allowed for the 32-bit system was also seven years. Could such projects based upon such a time-frame ever have succeeded in the face of the rapid progress taking place internationally? In the event, to accommodate ECIL as the national champion of computers, computerization was held back. The Automation Committee Report and H.N. Mukerjee's Public Accounts Committee hearings came

Technology and Applications Development

227

in handy in order to hold back the import of computers, which were considered expensive items by the DOE. In a similar way, BEL s programme in the field of peripherals was not subject to any critical review with regard to its time-frame or even its usefulness in view of the DOE's inability to standardize. In any case, of a total outlay on technology development of Rs. 16 crores spread over nearly ten years, Rs. 7 crores was allotted to a defence project, where the money could have been easily obtained from the Ministry of Defence. This left hardly Rs. 9 crores for other projects. The ECIL effort was allocated nearly Rs. 4 crores and the BEL effort about Rs. 1 crore (there were other allocations for BEL apart from the Rs. 45 lakhs indicated in the Table 10.1) and the balance distributed over a variety of small schemes. The massive funding in the case of ECIL was a step in the right direction — but the funding or the effort did not match the requirements. The causes of the eventual failure of the programme were never frankly analysed either by TDC or by the Electronics Com­ mission, and timely steps were not taken to avert a morale-shattering end to the ambitious national programme. Some 228 computers were built by ECIL (of TDC-12, TDC-16, TDC-32 varieties), but as has already been pointed out, the goal of national self-reliance was not achieved and ECIL had to seek technology from Norsk of Norway, which had started com­ puter operations about the same time as ECIL. A discussion of the difficulties faced by ECIL and an attempt to find a solution never came about. The DOE and the Electronics Commission blamed ECIL but made no attempt to discover the reasons or to take steps to set matters right. Usually the financial advisers associated with the different ministries perform a useful role and, very often, critically evaluate the usefulness of the expenditure from a practical angle. The IFA attached to the DOE is reported to have brought to the notice of the Member, Finance, Electronics Commission, in 1982 that the TDC/NRC action had until then been not one of origination but solely of response to the projects sub­ mitted by various establishments. The IFA felt that an approach paper, spelling out the objectives to be attained by 1990 was necessary; he wanted a detailed listing of the needed products/processes and tech­ nologies to be identified and a strategy outlined to fulfil the needs by undertaking projects with time-bound objectives. He recommended that technology should in general be purchased and that development should be undertaken only in areas where such technology purchase was not possible — such as in the area of defence. He had doubts about the actual benefits that had until then accrued to the industry through the various

228

India and the Computer

TDC projects and wanted a full cost-benefit analysis. As all the TDC programmes were designed to build up self-reliance and to avoid the import of technology, the IFA's views, in that they ques­ tioned the fundamental assumptions, should not have been brushed aside on the plea that the benefits of research work are not quantifiable. A goal/mission-oriented approach is the only way to check wayward R&D expenditure. S.R. Vijayakar, Secretary, DOE, appears to have also felt the same. He took action in 1986 to decentralize the TDC funding to the different Directors of the DOE so that some accountability for the results could be ensured. This decision was again reversed in 1988. Six working groups were constituted, with an apex body — reverting to the situation which prevailed earlier. There are eleven members of the Apex Body and eight to eleven members of each of the working groups. Such a large body, to disburse a sum of Rs. 3 to 4 crores per annum, appears wasteful of the time of so many senior scientists and executives, apart from the large expenditure likely to be incurred in the process. The government would do well re-examine the usefulness of the TDC set-up. The meagre allocation of resources without any time-bound definite programmes appears not to have led India anywhere. The entire computer technology development exercise in the period 1970-84 has to be con­ sidered a failure, though successes were seen in special areas, like nuclear reactor instrumentation, and in the case of some dedicated applications. A rhetoric-free appraisal of the realities of technology development, its costs and time-frames, is essential. Industrial development has proceeded on the basis of the import of CKD/SKD kits, and of foreign technical collaborations. Indigenous government-sponsored R&D has been more or less irrelevant for industrial progress in the computer field in India.

REGIONAL COMPUTER CENTRES The concept of Regional Computer Centres, (RCCs) conceived by the Electronics Committee (Bhabha Committee) and later supported by the Narasimhan Committee in 1968, came into being almost a decade later. Though M.G.K. Mcnon was committed to the idea, the mechanics of obtaining funds and co-ordinating the plans with different agencies and organizations took time. The Centres only became operational during 1977-9. The Calcutta Centre was equipped with a B-6738 Burroughs Computer which became operational in 1977. According to the balance sheet of the

Technology and Applications Development

229

regional computer centre dated 31 March 1987, the DOE had contributed Rs. 3.16 crores, the West Bengal Government Rs. 53 lakhs, and UGC about Rs.83 lakhs. The centre earned a revenue of about Rs. 60 lakhs in 1986-7 and was able to meet its recurring expenditure. Apart from the education and training activity, the centre also provides consultancy services and undertakes software development. The Burroughs System has been recently replaced by a CDC 840 financed by the DOE. The CAG has criticized this DOE-funded purchase as the site at which to install the computer had not been decided on at the time of purchase. The delays cost the DOE a further sum of Rs. 142 lakhs towards custom duties, etc.6 It has been reported that the utilization of the new Cyber machine has been very low and the management of the RCC unable to cultivate a market. CMC Ltd., has now stepped in to run the RCC at Calcutta. CMC's own performance in using the Indonet mainframes as data centres has not been a spectacular success. CMC Ltd., in their Annual Report, 1988-9, noted that another CDC 180/840A computer has been installed at the Jadhavpur Regional Com­ puter Centre. Both these sales represent a loss of business to ECIL, of course, who could have supplied the Medha 930 computers. In view of the poor utilization of the RCC computer, the import of these machines points to the failure of the DOE in assessing the needs; sales pressure of the agents and an urgency to exh lust the budget provisions were, perhaps, the compelling reasons. The Chandigarh centre acquired a DEC 2050 computer system. The total grant-in-aid received by the centre up to March 1987 was Rs. 2.09 crores, which was comprised of contributions from the DOE (Rs. 1.71 crores), the Punjab Government (Rs. 20 lakhs), and the Punjab Engineer­ ing College (Rs. 18 lakhs). A further grant of Rs. 75 lakhs was given by the DOE in 1986-7 to finance the purchase of the new DEC VAX computer. The DEC 2050 computer system was delivered on 16 April 1979, and the acceptance test completed on 29 May 1979. The RCC was registered in March 1979 and the computer system commissioned in May 1979. The 1981-2 Annual Report of the DOE obliquely referred to some administrative difficulties then being faced by the centre. It noted:7 This Centre is under a critical phase of re-structuring. In October 1981, a nominee of the DOE was appointed as administrator to take administrative and financial charge of the Centre. Subsequently the Centre has been put in charge of an officiating director.

230

India and the Computer

Serving as a data processing centre for various government organiza­ tions in Punjab and Haryana, the Chandigarh centre had been able to realize an income of about Rs. 1.17 crores in 1986-7. This income attracted income tax — a problem which the centre has yet to resolve. The RCC was able to report a surplus of about Rs. 7 lakhs in 1986-7, after allowing for an income-tax payment of Rs. 29.22 lakhs. The Executive Council, in February 1986, decided to replace the DEC system with a new one, a DEC VAX 8650 system (DOE-funded), ac­ quired through Hinditron Computers The Director of the Chandigarh Centre, D.R. Taneja, made the fol­ lowing points in reply to queries:8 We have not obtained any software from NCST (NCSDCT). However, some of our staff members attended some courses in NCST and also at I1T Kanpur, who are having a DEC 10 system. We had some interaction with 11T Kanpur, and we are trying to arrange some software for VAX 8650 from NIC. ... The software which we are trying to obtain from NIC will be paid by us. We have not taken up any research project for advancement of computer science because the Centre has to function on a self-sustaining basis. ... The software developed by RCC cannot be termed as routine packages. Those who are developing application software can appreciate the effort that goes into development and maintenance of software. ... The Centre has not published any paper in Indian/Foreign journals as our foremost concern is to earn enough funds to keep the Centre running.

The study team on computers noted:9 In the past it has been noticed that the usage pattern of these Centres are not in line with the objectives laid down for such Centres.

It was also mentioned that the DOE had constituted a committee in September 1980 with V. Rajaraman as Chairman to give a new direction to such centres. This committee's report was either not accepted by the DOE, or the committee did not complete its task, as no report has been recorded. The regional centres have mainly proved useful in providing data processing facilities and some consultancy work and development of software for specific projects. The concept that the centres would be used by Indian software engineers for software export has not proved to be the case — no such activity has been reported by the RCCs. With the liberalization of imports from 1983-4 and the increased pressure from the selling agents of foreign companies, several imported systems came into the country for use in public sector organizations, Institutes of Management, and other autonomous bodies. The PC wave

Technology and Applications Development

231

also provided alternative solutions for many data processing applications. Computerization got a big boost and the relevance of RCCs has therefore correspondingly diminished. The upgradation of the computer systems in the RCCs will require a large infusion of funds.The RCCs may also have to generate its own funds just in order to survive, though their manage­ ment culture may not currently be conducive to this kind of approach, although the NCSDCT did manage to generate its own funds for the acquisition of new computer systems. As the revenues of the RCCs are now subject to income tax, the surpluses of the order required to make purchases of large computer systems may indeed be difficult to come by. Currently, the DOE does not appear to have any plans to set up more RCCs in other areas, as the NIC is itself expanding setting up a national computer grid. Funds may, therefore, be difficult to find for new regional centres. As Rajaraman noted, it would be better for the RCCs to be incorporated into universities instead of their having a separate 'society' status. The universities would be able to make better use of them in their various educational and research programmes, preventing the RCCs from ending up as mere data processing centres. THE NATIONAL CENTRE FOR SOFTWARE DEVELOPMENT AND COMPUTING TECHNIQUES The Electronics Commission, realizing the importance of software development, decided to seek UNDP assistance for setting up a National Centre for Software Development and Computing Techniques (NCSDCT) at Bombay to be managed by the TIFR, and headed by R. Narasimhan of TIFR. The UNDP supported the centre with an initial contribution of US $ 2.639 million. The computer centre became opera­ tional in the last quarter of 1975. The facilities at the centre comprised a large DEC 1077 (dual-proces­ sor) computer system with an interactive computing capability and asynchronous and synchronous communications processing capabilities. The centre also had a PDP 11/40 based Vector General interactive graphics facility. The 1975-6 Annual Report of the DOE defined the centre's objectives, which included the development of generalized software packages.10 But the NCSDCT thought differently. The summary of the activities of the NCSDCT noted:11 The basic objective of NCSDCT is to contribute to the growth of computer technology and computation practices in India ... concerning itself with

232

India and the Computer

technology development and dissemination but not just in terms of developing products and distributing them. Playing a role which is restricted neither to that of a software house nor to that of a computer science department, NCSDCT has sought to promote technological innovation through examples and instruction. With generous assistance from the Electronics Commission, NCSDCT has been working on the development of software tools and techniques, and also on the realization of specific systems in three major task environments of immediate national relevance. These are: 1. Software Engineering: The development of large scale system software like language processors and operating systems; 2. Computer networking: The development of software for computer and computer communication; 3. The development of graphics-oriented Computer-Aided Design . (emphasis added)

A comparison of the centre's role as envisaged by the DOE and that which the NCSDCT had set for itself shows that whereas the DOE wanted some software and software packages to come out of the efforts of the centre, the centre felt that their role was mainly promotional — training, teaching, consulting, and generally academic in approach. During the period 1974-9, the NCSDCT developed various pieces of software for use on the DEC 10 and the TDC-316. Their effort on behalf of ECIL for the TDC-316 system was part of the national computer programme. As the TDC-316 computer did not become popular in the market and was being rapidly superseded by other imported/assembled computers, their software effort for the TDC-316 was unfortunately not of lasting significance. DEC 10 packages must have been of assistance to NCSDCT in throwing open the DEC 10 system for exploitation by both trade and industry in Bombay. Other DEC users, like the RCCs at Chandigarh and Roorkee, or the IIT Kanpur, do not seem to have benefited from the work at the NCSDCT. A profile of system usage compiled by the NCSDCT indicates that nearly 30 per cent of computer time was used for commercial work yielding revenues for the NCSDCT. The centre, therefore, seems to have become a data processing facility for the Bombay area, though the NCSDCT claimed that the centre was not generally used in this way. The staff of the NCSDCT and the T1FR were able to act as consultants, receiving payment for their work. The availability of the DEC 10 system combined with the NCSDCT's expertise must have attracted a lot of clientele. To an extent, the Centre also became a teaching set-up with courses

Technology and Applications Development

233

being run on chosen topics, with visiting professors from abroad lecturing there. After M.G.K. Menon relinquished his appointment as the Chairman of the Electronics Commission, the link between the DOE and the TIFR (through Menon's previous close connection with the TIFR, and his long, personal association with Narasimhan) came to an end. The new Chair­ man of the Electronics Commission, Nag, wanted, perhaps, to exercise a more direct control over the affairs of the NCSDCT. The Review Com­ mittee (the Sondhi Committee; Nag, a member) recommended in 1979 that the NCSDCT should be hived off from TIFR and placed directly under the DOE.12 Though the recommendation of the Review Committee was accepted by the government, the status quo continued until 1985. The period 1979-86 saw the NCSDCT continue its working in the set pattern. The UNDP came forward with further assistance for organizing workshops in specialized areas of software technology and for conferences in theoreti­ cal and conceptual aspects of software technology in order to strengthen the academic base of software engineering activities in India. The NCSDCT organized these workshops and brought leading computer scientists to India for academic interaction. Out of its own resources, the NCSDCT acquired a DEC 2020 System and installed this at the Air India building, Nariman Point, Bombay. The PDP 11/40 system was also moved to the Air India building. A Unix-based dual processor HP-9040 system was acquired and com­ missioned in August 1985. A circuit switch facilitating access to the three machines (HP-9040, DEC 10, and DEC 20) from any terminal connected to the switch was developed by the NCSDCT and remote access to the PDP 11 from the HP-9040 also made possible. The NCSDCT did not want to become goal-oriented or missionoriented and to tic their work to national projects. The role they defined for themselves allowed them to work in various areas of interest to the staff members but did not lead the centre to develop software of general relevance to the nation as a whole. No operating systems or other pack­ ages of national or international significance were produced. Their inter­ ests were essentially academic, and to a certain extent, dictated by their clientele in the Bombay area who came to them for consultations. The output of NCSDCT papers and some of their other activities are listed in Table 10.3. The DOE constituted a further committee in 1983 to review the NCSDCT, with A. Krishnaswamy, IAS (Rctd.), as Chairman, Nitin

234

India and the Computer Table 10.3 NCSDCT 'Paper' Output Count 1974-9 1979-80 1980-1

Technical reports (local) 38 Publications in journals Indian/Foreign Papers presented in conferences Books published/edited Doctoral theses Theses guided Theses submitted Participation in confe­ rences outside India Conference proceedings Popular articles Participation in conferences and workshops

1981-2 1982-3 1983-4 1984-5

15

21

5

11

20

5

11

4

4

6

10

9

8 1 1 3

14 1 1

5 1

10 1 3 8

9

8

-

-

-

-

-

-

7 1

-

12

NOTE: The Annual Report, 1985-6, also recorded the publication of three papers in international journals, three technical reports, six contributions to conference proceed­ ings, and the publication of a book by K. T. Sridhar, Programming with Pascal. SOURCE: Compiled from the Annual Reports of the NCSDCT.

Dcsai, Adviser, Planning Commission, and J.C Khurana of Bharat Petroleum, as members. The committee was asked to present a clear vision for the NCSDCT's future role and its mode of operation, retaining all that was relevant from the past and recommending possible new directions. The report of the Krishnaswamy Committee contains no review of the work done till then or of its national relevance. Perhaps the Chairman and the other two non-NCSDCT members were not inclined to closely ques­ tion these aspects or were overawed by the aura surrounding an estab­ lishment so closely linked to the prestigious TIFR. The Policy Review Committee recommended that the NCSDCT should be a centre of excel­ lence in software technology and related computer science with a com­ mitment to achievements having economic significance. Regarding the NCSDCT's role, the policy committee noted: 1.

It will be R&D oriented, using its judgement and perceptions to choose

activities at the emerging frontiers of technology.

Technology and Applications Development

235

2. In its practical activities, NCSDCT will emphasize software production more than activities such as consulting, turnkey projects, and providing facilities to other groups. 3. The criteria for performance of NCSDCT will not give weight to quantities such as the turnover, size of market segment captured etc., but to coverage of significant disciplines in the frontiers of technology, excellence of its R&D work, national and international impact of its R&D in influencing the work of other groups through the concepts and tools it develops, and through its publications, etc. 4. NCSDCT should strive to be 'front runners of the pack as far as technology is concerned, leaving large scale exploitation of the market for services to others. It should drop areas as they become well understood, and pick up new areas which offer R&D challenges.(emphasis added)

The Electronics Commission was to consider the report of the Policy Review Committee and proposal for a plan of action prepared by the NCSDCT at its seventy-second meeting, originally scheduled for 28 April 1984 but then postponed till 21 July 1984. Meanwhile, P.P. Gupta relinquished charge as Secretary, DOE, on 25 May 1984. The presenta­ tion of the report was again postponed till the seventy-third meeting on 17 August 1984, when the proposal was approved. On 21 March 1985, a new society named the 'National Centre for Software Technology' was registered in Bombay under the Societies Registration Act of 1860. The registration of this new society formalized the spinning off of the NCSDCT from the T1FR and established it as an independent institution — the National Centre for Software Technology (NCST) — a welcome simplification of the tongue-twisting acronym NCSDCT. The budget support from the DOE, prior to 1980, was small — in 1975-6, Rs. 20.47 lakhs; in 1978-9, Rs. 8.10 lakhs; in 1979-80, Rs. 2.47 lakhs. With the hiving off of the NCSDCT, the DOE support increased. The Sixth Plan expenditure amounted to Rs. 2.67 crores and the Seventh Plan expenditure Rs. 3.69 crores. The proposed outlay for 1990-1 was Rs. 2 crores. In their anxiety to release funds and to utilize the budget allocations, the DOE invited a comment from the CAG about premature release of Rs. 144 lakhs to the NCST.13 The CAG noted that though the computer was purchased in 1987, land for the advanced training facility had not been acquired by 31 March 1988, the advanced training facility had not materialized, and the NCST had Rs. 74.71 lakhs of unutilized funds in its hands. The budget allocation sought for the Eighth Plan is Rs. 18.5 crores.

236

India and the Computer

The NCST is managed by a Governing Council heavily weighted with the DOE representation. The 1985-6 Annual Report — the first report of NCST — noted: NCST is the latest contribution of TIFR to the nation. Its main strengths are neither its new computer systems nor its new office premises in Juhu. Valuable as they are, the heart of NCST is the scientific and administrative tradition, and the value system its staff have inherited from their R&D careers at TIFR.

Narasimhan relinquished charge of the NCST on 13 October 1987, S. Ramani taking over charge as director. The NCSDCT, and its later offshoot the NCST, came into being because of the efforts of M.G.K. Menon and R. Narasimhan. They were able to win a generous grant from the UNDP to the extent of about US $ 3 million and build up a team of dedicated software specialists. The grants given to the centre by the DOE till 1980 only totalled Rs. 31.04 lakhs — the NCSDCT was largely a self-supporting organization till 1985. NCSDCT supported the national computer effort at ECIL but the interac­ tion between ECIL, the DAE, the Electronics Commission and the TIFR did not result in any enduring hardware or software. Had the TDC-316 and 332 systems caught on, and provided the necessary computing power for the national needs, the NCSDCT may have been in a different position. Neither did the NCSDCT record breakthroughs in the areas of languages or software. The NCST is gradually becoming an academic teaching institution following the typical pattern in India. To become a centre of excellence, it will have to demonstrate work of an excellent standard attracting the attention of the world computer science com­ munity cither that, or produce nationally relevant software with widespread applications.

THE NATIONAL INFORMATICS CENTRE Both the Indian Statistical Institute, Calcutta, and the Department of Statistics, Cabinet Secretariat, New Delhi, introduced the use of com­ puters for government work early on. In 1967, the government acquired ten Honeywell computers. A computer centre was set up in the Depart­ ment of Statistics with three of the Honeywell computers in order to provide data processing facilities to all the government organizations located in and around Delhi. The remaining computers were deployed at other important centres for similar purposes. The Computer Centre, New Delhi, had ambitions to set up a central facility and in a note to the PAC

Technology and Applications Development

237

it had stressed the advantages of a large computer for government needs.15 In the tussle for supremacy in the area of government computerization, the Electronics Commission came out top. The Finance Secretary and others were no match for Menon's powerful eloquence. The PAC, with H.N. Mukcrjcc as Chairman, advocated a leading role for the Electronics Commission and the DOE concerning all aspects of the introduction of computers in India, in monitoring their use, and in providing technical leadership. At that point of time there were hardly two or three ex­ perienced computer scientists in the DOE — Mcnon was banking on his associates at the TIFR to help him out. Thereafter, the role of the Department of Statistics and ISI, Calcutta in the computerization of government work was relegated to the back­ waters. The Planning Commission had its own computer acquired from IBM under a Ford Foundation grant, but they were persuaded to let the DOE take the lead in building up a national data base. The Electronics Commission and the DOE approached the UNDP for assistance for establishing a large computer centre at Delhi for building up national data bases, developing the methodologies for utilizing these, and for defining the various options and paths in decision-making at the national level. The original intention was to locate the National Computer Centre in Delhi at the Jawaharlal Nehru University (JNU) for databank management and the development of methods to handle and make op­ timal use of large amounts of information, particularly in the context of national planning. A UNDP mission visited Delhi in March 1975 to study the proposal. The UNDP agreed to fund the National Informatics Centre to the extent of US $ 4.4 million for the purchase of a large computer system (costing approximately US $ 3.3 million), other hardware, train­ ing, the services of experts, etc. The 1976-7 Annual Report of the DOE, noted:16 In view of this, action has been taken to proceed with the preliminary work relating to the National Informatics Centre so that when UNDP financial assistance becomes available and the major hardware is commissioned (in 1978 on the basis of current information), the Centre can get on with its work on an expeditious (sic) basis. NIC is a plan project administered by the Information, Planning and Analysis Group (1PAG) of the Electronics Commission in its initial stages. It may be noted that it was decided that the computer centre was not to be a part of JNU after all.

India and the Computer

238

The advisory council set up for the NIC in 1976-7 did not include any member from the Department of Statistics, the Cabinet Secretariat or from the Indian Statistical Institute, Calcutta, though these organizations had until then been handling government data. The reason for their exclusion from the NIC is not clear. The Controller-General of Accounts, Ministry of Finance, was a member of the council, but his experience in systems design of government data bases may not have been adequate for the job. It is also not clear whether there was any dialogue with the Department of Statistics or with ISI in formulating data base systems for the Planning Commission or for other government departments. The long term objective of the National Informatics Centre, as ap­ proved by the Planning Commission, Ministry of Finance, and the Electronics Commission, was to establish a system for the provision of detailed information for the government, the ministries, and other related government agencies, so as to assist them in making decisions relating to' the country's plans for economic and social development and in programme implementation. 1 7 The following information systems were planned: 1.

An agricultural information system.

2.

A small-scale industries information system.

3.

An energy information system.

4.

An education and manpower information system.

5.

A socio-economic index information system.

6.

A finance information system.

7.

A construction and transport information system.

8.

A trade and media information system.

9.

An industrial information system.

10.

A government archival information management group.

Many others were added to the list from time to time. The task was highly ambitious and would require an enormous amount of interaction and co-ordination between those developing the data bases and those who would use it to derive significant, reliable information with decision­ making relevance. The UNDP-financcd Control Data Corporation computer, a Cybcr170/730 system, was commissioned in August 1980. The system con­ sisted of dual CPU, one 96K-bytcs (60-bit word) memory, 2.5 billion bytes of disk storage, 9 magnetic tape drives, 2 line printers, 2 card readers, one batch terminal, 3 network processing units, 30 terminals and one system console. The NIC had acquired the biggest computer in the country, but from the initial conception to commissioning had taken eight years.

Technology and Applications Development

239

Table 10.4 NIC Financial Outlays (Rs. lakhs) Year 1974-6 1976-7 1977-8 1978-9 1979-80

Actuals 41.65 49.41 96.50 129.07 203.54 520.17

Total up to 1979-80 1980-1 1981-2 1982-3 1983-4 1984-5

199.95 145.28 191.73 179.53 2,542.30 3,258.79

Total Sixth Plan 1985-6 1986-7 1987-8 1988-9 1989-90 1990-1

2,924.55 2,906.89 3,500.00 2,554.00 2,031.00 * 1,370.00 **

Total 1985-91

15,286.44

Total

19,065.40

* Revised estimate; ** Budget estimate. SOURCE: Annual Budget Estimates of the DOE. Apart from the UNDP grant, the expenditure on the NIC as presented in the budget ctimatcs of the DOE can be seen in Table 10.4. Up to the end of 1983-4, only Rs. 12.36 crorcs had been spent on the NIC. The spurt in the expenditure from 1984-5 followed Rajiv Gandhi becoming the PM. Regional NIC cells were established at Bombay, Calcutta and Madras for the collection of information. The work of the NIC got side-tracked in 1982 when it was entrusted with the task of developing a 'results information system' for the Asian games. Again, the Seventh Non-Aligned meet (NAM) proved another

240

India and the Computer

diversion as the NIC was required to process documents for it. The NIC also put up an electronic mail system for media personnel seeking inter­ views with the heads of delegations. These two tasks, not really related to the objectives of the NIC, provided the Director, Seshagiri, with close contacts with Rajiv Gandhi and the Prime Minister's Secretariat. A further boost to the NIC, the acquisition of additional computer systems and the setting up of a national network was finally realized because of the involvement of top political personalities who were swayed by the Asian games and the NAM conference successes. In the Status Report of NIC (1983), Seshagiri was all praise for the NIC.

The euphoria over the of Asian games, NAM, and his association

with top personalities in the government because of these two events, seems to have blurred Seshagiri's vision of the tasks ahead in setting up an information network of real utility and use. One would have expected to see the financial outlays on the NIC and an assessment of it workings in the status report, but these were unfortunately lacking. Seshagiri told newsmen in 1985 that the government computer network at the centre was to be extended to all the states and union territories under a decision taken by the Prime Minister, Rajiv Gandhi.19 Seshagiri said that ultimately all the four-hundred odd district centres would be inter­ linked. According to Seshagiri, the use of computers for the collection and dissemination of data would reduce the delays in decision making and minimize the need for

ad hoc decisions. He mentioned that in the first

phase, four supercomputers would be installed at Delhi, Hyderabad, Pune and one in Bihar or Orissa. The computers were to be obtained from NEC, Japan. The then Karnataka Chief Minister, Ramakrishna Hegde, (made the Deputy Chairman of the Planning Commission in 1990 by the Janata Dal minority government; he later resigned) noted:20 The idea of acquiring a supercomputer to control and monitor everything from Delhi is totally undesirable.... The introduction of supercomputer is an indirect attempt to divert the attention of the people from economic and social issues like unemployment, inflation, poverty and illiteracy ... the Centre could utilize the supercomputer for 'intelligence purposes' as well. Hegde, however, clarified that he was not against computerization as such, and noted that Karnataka had already introduced computers in Karwar district. He felt that the centre could secure data from the states and feed it into its computer instead of having its own supercomputer network throughout the country.

Technology and Applications Development

241

Hegde appears to have later relented, and provided accommodation for the NIC in the Visvesaraya Building in Bangalore and an understanding has been arrived at for the NIC to operate in Karnataka. NEC supercomputers were installed at Delhi, Pune, Bhubancshwar and Hyderabad, and slowly taken into use. Apart from setting up the computer network environment, different data bases had to be designed and sup­ ported by application programmes, documentation, and data entry forms. To carry this out on a nation-wide basis is a mammoth task, especially if one is to ensure the usefulness of the data bases and their credibility for the task. Regional chauvinism, regional language support, and local apathy had to be countered. The NIC claims to have developed over 300 Information Systems for some 40 ministries, departments and associated organizations of the Central Government. The Central Government Informatics has been segmented into nine major sectors: apex organizations, agriculture and water, industry services, energy, science and technology, finance and commerce, human resources, and security. The NIC reports that over 400 districts have now been connected to NICNET and that over 1200 professional users belonging to govern­ ment/associated organizations are using its facilities. The NIC District Information System (DISNIC) (covering 27 sectors) is now being imple­ mented in many districts, enabling quick access to information by the state and central governments. The NIC has also set up general information service terminals (GIST) which enable the public to access some of the general data bases. GIST (Version 1.0) is also available over NICNET. All the modules are menu driven, user friendly and self-explanatory. GIST could prove very useful, if the information content is reliable and up-to-date. The Informatic services of NIC are supported by a staff of 745, of which 570 are computer professionals. The revenue expenditure of the NIC was Rs. 22.50 crorcs in 1989-90. This is projected to increase to Rs. 26.50 crorcs in 1990-1. In February 1988, the government rightly decided to place the NIC under the Planning Commission. A review committee was also set up to look into the affairs of the NIC. It may be necessary, for instance, to consider merging the activities of the other computer bureaus of the central government with the NIC so as to ensure the better utilization of facilities and the avoidance of a duplication of effort in data gathering. A minor competing programme, CRISP (Computerized Rural Informa­ tion Project), has also made progress in co-operation with the state

242

India and the Computer

governments.21 There appears to be some overlap between the two and the Planning Commission would do well to sort matters out before an unnecessary duplication of effort arises. The large scale use of computers brings with its own problems — problems of concerning the manipulation and falsification of data. A classic example quoted by Daniel Bell and used by Weizenbaum relates to the command and control system in operation during the Vietnam war.22 Admiral Moorer, then Chairman of the Joint Chiefs of Staff, testified to the US Senate Armed Services Committee that specially programmed computers in the field systematically lied to the Pentagon's computers with respect to the secret bombing of Cambodia ... Computers in the field were programmed to automatically convert the geographical co-ordinates of targets struck by US planes in Cambodia to co-ordinates of 'legitimate targets' in Vietnam ... a text-book example of Orwell's Ministry of Truth. India should beware. Centre-state relations are not harmonious. The possible falsification of data bases for political advantage cannot be ruled out. Inaccurate and inadequate economic indices could cause incorrect decisions to be taken. It may be perhaps wise to separate the NIC from direct government control and set it up as an autonomous organization to insulate it from political interference. But this could deny the NIC access to government files, making information gathering more difficult. The move to grant constitutional status to the Planning Commission may point to a way forward through the Commission itself is increasingly becoming a political instrument of the party in power. It should also be noted in this context that the speed with which decisions are made and the making of correct decisions cannot be corre­ lated with the ready availability of large volumes of data. William Plowdcn, Director-General of the Royal Institute of Public Administra­ tion, London, is reported to have told the Delhi staff reporter of The Hindu that the age-old concepts and principles of public administration will not change grcatcly because of the technological revolution.23 He felt that the computer cannot change the relationship between the politicians and the administrators — the two key groups — as it docs not effect the skill of administration. At best, the computer could do away with 'low grade' clerical work. He further noted that man's genius and sincerity of purpose in solving the complex problems of society will have to remain a human endeavour. The computerization of the Government of India will require a major

Technology and Applications Development

243

co-ordinated effort. The various departments of the government have computerization plans of their own. Ensuring the co-ordination of all the activities including that of the NIC, with the ultimate aim of networking, is a major task. The Seventh Plan outlay on NIC/NICNET was about Rs. 139 crores. An expansion to the 'block' levels is being considered under the Eighth Plan which would result in an outlay of Rs. 360 crores. The annual revenue expenses have risen to over Rs. 27 crores. A review by the Estimates Committee of Parliament may therefore be useful. The NIC was conceived of some fifteen years ago, and a UNDP-funded CDC computer installed ten years ago. A substantial capital expenditure, of over Rs. 190 crores, has already been incurred. Though Seshagiri has applied himself assiduously to the task with some imagination, and has managed to set up a vast network, its ultimate utility will depend upon the accuracy and timeliness of the information and the government's willing­ ness to use that information in sensible ways. The NIC is the single most ambitious programme in the area of application of computer technology for national development.

THE CLASS PROGRAMME The Class programme (Computer Literacy And Studies in Schools) started with a proposal considered by the Electronics Commission for a pilot project in secondary schools planned for the end of 1983. The objectives of the DOE plan were: 1. To provide students with a broad understanding of computers and their use. 2. To provide hands-on experience. 3. To familiarize the students with the range of computer applications in all walks of human activity and the computer's potential as a control and information processing tool. 4. To demystify computers and to develop a degree of ease and familiarity with computers which would be conducive to develop­ ing individual creativity in identifying and developing applications relevant to their immediate environment. The emphasis of the programme was on the development of manipula­ tive skills rather than on teaching the principles of computer science. The proposal was to introduce the programme from the top, i.e. the programme would first be introduced at the secondary/higher secondary schools and colleges, and would then be extended to middle and primary schools. The reason for starting the programme at the secondary school

244

India and the Computer

level was that those students would enter the nation's work-force sooner than the others and should, therefore, be given priority. It was hoped that a further programme for computer literacy at the middle and primary school levels would be taken up later. As the country has about 5,500 colleges, 55,000 secondary and higher secondary schools, and about 620,000 middle and primary schools, it was obviously not possible to launch the programme at all these centres. It was, therefore, proposed that in the final year of the Sixth Plan (i.e. 1984-5) to launch a pilot project at a limited number of institutions of a similar character in terms of the language of instruction, course cur­ riculum and administration. It was thought that the Central Schools would meet this criterion and would so prove suitable for the initial effort. There was also an idea mooted that after establishing the computer literacy/computer education programmes in the schools, that the com­ puter should also be used for instruction purposes. The idea was to motivate learning through the use of animation and graphics, and to convey complex concepts by the process of simulation. The intention was not to displace or replace teachers but to enhance the ability of the teachers to educate more effectively. Plans to develop CAI (ComputerAided Instruction) software were to be launched separately. The original plan was to cover 500 schools but the Prime Minister, to whom the project had been submitted in advance, reduced the number to 250 schools on account of resource constraints. The extent of coverage in the Seventh Plan was to be decided after assessing the lessons of the pilot project. The Prime Minister added, 'Let us make as quick a start as possible.' The decision to go ahead was taken by the Chairman, Electronics Commission, and the Secretary, DOE, in consultation with the Prime Minister. The planners at the DOE seem to have been rather optimistic. They stated: It is not unrealistic to assume that the computer literacy/computer education programme can be implemented country-wide during the Seventh Plan. An investment of Rs. 300 crores is likely to be required for covering all institutions. However, Dhawan was critical of the whole approach, that it had not been well thought out or well planned. His reported observations were: 1. Among the aims and components of the Pilot Project were the definition of the details of the educational objectives' and the design of the curriculum and preparation of course materials. It appeared surprising to Dhawan that such an important project was being attempted without

Technology and Applications Development

245

having established a major part of the curriculum and materials. He was unable to see how the project could define this aspect. No doubt, the pilot project, being an active real-life project would, through the experience gained, provide data for making changes and improvements — but it could not generate the curriculum, etc. There was no information on the status of teaching materials, etc. 2. The Teachers' Training Programme should have logically preceded the Pilot Project. Otherwise the whole project would be in jeopardy. 3. He asked whether the number of hours of lectures and practical sessions be in addition to the normal load of the students in schools. If so, the load would be such that either their normal studies or the 'computer literacy' project would suffer. There was no indication of how this conflict would be resolved or how it would be handled by the school bodies. 4. There was no systematic and concrete evaluation criteria or methodology for assessing the results of the pilot project. If this is not done beforehand, it later becomes solely a matter of debate and of pure subjective judgement. The evaluation criteria and methodology should have been laid down, and the agencies and teams identified before the project began. 5. The school distribution, with 245 schools out of 250 being govern­ ment schools, did not appear to be representative of the secondary school system. 6. Dhawan also felt that the fact that the pilot project would complete­ ly be based on imported computer systems was not a very happy situation. That a workshop organized by the National Council for Education, Research and Training (NCERT) had discussed the issues and formulated a curriculum was sufficient for the DOE, however. At the pilot stage, the computer literacy project was to be a voluntary, a non-cxaminable subject and part of what is classified as socially useful productive work (SUPW), for which time slots arc supposed to exist in the school curriculum. The NCERT, perhaps the most knowledgeable institution in the country in the field of education, had reservations about the scheme. They were muted in criticizing the scheme in view of the PM's well-known passion for computers. The Director of NCERT, in his address at the National Workshop held at NCERT on 26-7 March 1984 said:24 More than 35 per cent of our primary schools are managed by single teachers and

India and the Computer

246

a large number of secondary and middle schools are yet to provide basic facilities for teaching. In view of this background, one of our major concerns in curriculum development has been to ensure equality of opportunity to children of diverse socio-economic backgrounds while promoting innovative measures for universalization of elementary education and the improvement of the general quality of school education. It has also been observed in the recent past that the quality of science education could not be improved by investing on laboratory facilities alone. Very often such facilities remain unutilized or underutilized in the absence of proper motivation on the part of school administrations and science teachers.... There is also a growing apprehension among a group of educational philosophers that an excessive interaction with and dependence on the computer in the early stage of education may promote a highly structured and linear thinking among children, which may adversely affect their emotive development and holistic vision. ... Student motivation to participate in any optional course depends either on its relevance to the terminal examination or its recognition in the job market. ... Developing a computer-literate school is a complex process and there is an urgent need to systematically plan every step for the implementation of our pilot project so that appropriate steps could be taken immediately to overcome the known practical problems of failure. The then British Prime Minister, Mrs Thatcher, during her visit to India had agreed to give a grant for the Class programme based as it was upon the BBC microcomputer. Under a bilateral Technical Co-operation Aid arrangement signed on 11 January 1984, the UK government agreed to provide 900 computer systems, including software, for the Class project. Some 170 systems and the associated software were received in the second week of June 1984 for the conduct of a Teacher Training Programme at 17 Resource Centres. The UK government aid package amounted to £ 1.3 million and consisted of hardware, software and training components. Considering that the Indian business environment mainly use MS-DOS based PCs, a few hours on the BBC micro with untrained, uncnthusiastic teachers could only result in the waste of scarce resources. The task of evaluating the programme was assigned to the Indian Space Research Organization, Ahmcdabad, who undertook a study and sub­ mitted two reports — one covering 'dcmystification' and the other cover­ ing the 'impact on students'. The major findings of these studies were as follows: 1.

The Class project has three active partners for the project execution

— the DOE, the NCERT, the Department of Education. It has two kinds of institutions on the receiving side — the Kcndriya Vidyalayas and the

Technology and Applications Development

247

State schools, the servicing of the computers being carried out by CMC and the Resource Centres. The complex formalities and protocols and the plurality of agencies that have to be dealt with at each stage and the misunderstandings of the very nature and requirements of the computers all contributed to an inordinate delay in installation and commissioning. 2. The schools invariably selected the senior most science teachers to be in charge of the project and this led to unwilling teachers being recruited — teachers on the verge of retirement. After selection and training, more than 50 per cent of the teachers withdrew from the project. In the case of the Kendriya Vidyalayas, the transfer of computer-trained teachers depleted the overall number available for teaching. Some teachers were not serious about the training programme and the trainers of the (Resource Centre) appeared dissatisfied with the trainees. Also, because of inadequate computer facilities, many teachers, especially female teachers, could not get enough hands-on experience. Practical experience could not be adequately provided and several packages were not even tried out by the computer teachers. On the whole the teachers seemed to be dissatisfied with the various training programmes given to them, and the Class project had to operate with the added difficulty of inadequately trained computer teachers. Many of the teachers also had difficulties in following training in the English language. 3. Incorporating computer literacy in the existing curriculum has proved difficult because the subject is not examinable and is considered a mere hobby. The report suggests that computer learning should be made a regular subject with full-time teachers and the marks should be added to the aggregate for a pass. (The report ignored the financial constraints involved.) 4. The students generally are thrilled and excited about the prospect of learning computers. However, most of the students have difficulty in following English commands and programs. It has been suggested by the teachers and others that if the commands were in the regional languages, it would facilitate computer learning. It would appear that at the school level, regional language computer programmes would be required for 'dcmystification' and learning. 5. The report notes that only a few of the 1500 students were 'demystified'. The main reasons for this were that very few computer classes were held, proper theory and practical classes were not taken by the computer teachers and individual hands-on practice was difficult to come by due to the large batches of students handling each computer. The use of English created a barrier in most state schools.

248

India and the Computer

Even in the United States there are two schools of thought regarding computers in the class room. These views were highlighted in an article by Alison B. Bass.25 The developer of the computer language for children Logo, Seymour Papert, feels that the computer will help children learn faster and ap­ preciate the theoretical concepts better. On the other hand, the article noted that MIT researchers found it extremely difficult to measure in any systematic way what concepts the children were picking up. Papert explains that the reason why children do not learn mathematics is not because it is hard, but because it is not related to their experience. They cannot do anything with it that seems worth doing, and so it feels dead to them. He adds that in the real world people always learn by experience and he feels that we now have technology that children can use to make something they are interested in — whether it is pretty shapes with Logo or cars with Lego. He feels that the computer provides the children with a way of appropriating mathematical knowledge and in using it in a very personal way. Others like Weizcnbaum, Professor of Computer Science at MIT, differ. He says: Papert may, in fact, have found a better form of teaching. But what does that have to do with computers? If anything, an over-emphasis on computers obscures the real issue which is the need to fundamentally restructure the educational system and deal with the social problems that hinder children's natural urge to learn. Children may not be motivated in school because they are hungry or they have been abused at home or for any number of other reasons. Simply introducing computers avoids the question of why children may not be motivated in school. It converts a social problem into a technological problem and then tries to solve it by technical means. In that sense, the computer serves to inhibit the asking of important questions about the way our society raises and teaches its young. Weizcnbaum further argues that the kind of thinking learned through programming — logical, step-by-step analysis is only one limited varia­ tion of human thought applied to the very narrow domain of problem solving — but most human problems are not solved that way. The result would be that children will tend to discount their more important intuitive thought processes in favour of a more limited analytical approach. But many other experienced teachers in the USA favour the computer in imparting education. The 1989-90 Annual Report of the DOE mentioned that 2,330 schools had so far been covered under the Class programme. They were supported by 53 Resource Centres. The programme was expected to cover 3000

Technology and Applications Development

249

schools by the end of the Seventh Plan. A sum of only Rs. 20 lakhs was budgeted in 1989-90. The DOE estimate for 1990-1 was Rs. 1 crore. The Ministry of Human Resources Development (MHRD) has also made ad hoc allocations for the Class project, which is under their care. However, it is admitted that there are schools where computers are lying unopened in their boxes due to the lack of basic infrastructure. The CAG has held the NCERT responsible for the ineffective implementation of the CLASS project and the PM has also laid the blame on the NCERT.26 When P.V. Narasimha Rao was the Minister for Human Resource Development, he was reported to have held the view that the Class programme is elitist. He was more inclined towards providing the basic necessities, like blackboards, water, proper accommodation, etc., for the schools. The annual reports of the DOE have since begun to underplay the Class programme, and the Space Application Centre at Ahmedabad is exploring the feasibility of involving Non-Govcrnmcnt-Organizations (NGOs) in the programme. The question we have to face in India is whether computer learning (or dcmystification) is relevant in the context of the scale of the current problem of primary and secondary education and promotion of literacy. The manpower requirements for computer usage in India need not be linked to the provision of computers in such a diffuse manner in the schools. We just cannot afford such wasteful, purposeless effort. The problem of a shortage of trained computer manpower has to be addressed by specific, target-oriented, quantifiable programmes in training in­ stitutes. For the poor, specific schemes for talented children, which lead to employment are necessary. Richer private schools will find the money for computers anyway.

MANPOWER The need to develop manpower in order to meet the growing require­ ments of computerization in India began to receive attention from 1975 onwards when the Perspective Report prepared by the IPAG, DOE presented a forecast.27 However, as no computer policies were formu­ lated until the end of 1978, the matter received little attention, until Rajaraman undertook a study. His report formed the basis of the action then taken by the DOE.2* Rajaraman had suggested a teacher training programme, increasing the intake of the existing B.Tcch. programmes, initiating a new'Masters in Computer Applications' programme at ten educational institutions and a one-year crash course to quickly train thirty

India and the Computer

250

to forty students in each of ten metropolitan centres. He made various other suggestions for improving the availability of personnel for com­ puter and microprocessor application areas. The Department of Electronics set up a Standing Committee on Com­ puter Education with representatives from the Ministry of Education, the UGC, and the academic community. This committee drew up a phased five-year plan for providing computer facilities in educational institutions.2y The following computer requirements were envisaged for the period 1980-5 (sec Table 10.5). Table 10.5 Manpower Development Computer Requirements (1980-5) System Cost

Ministry/Organization

Up to Rs. 0.5

Rs. 0.5

million

million to

Above Rs. 2

2 million

million

20

20

2

24

22

4

University Grants Commission Ministry of Education

SOURCE: 'Report of the Panel on Computer Manpower Development', I PAG Journal, vol. 8, 2, New Delhi, November 1980, pp. 55-70.

The Annual Report of the DOE noted:30 The Standing Committee on Computer Education has identified that there is a great need to ensure commonality between the different systems in educational institutions. This would ensure portability of software, lower maintenance cost and exchange of educational material.

The Electronics Commission approved the Rajaraman Panel Report in March 1981, subject to the actual expenditure by the DOE during the Sixth Plan being limited to Rs. 5 crores.31 The programme was to be implemented by the DOE, the Ministry of Education and the UGC. Programmes were started in the area of teacher training in 1981-2 at the IITs in Delhi, Madras, Bombay and Kanpur and also at the Jadhavpur University. B.Tcch. and M.Tech. programmes in Computer Science were also launched at various universities, the IITs and the engineering col­ leges. A plan was worked out by the DOE and the UGC for launching a

Technology and Applications Development

251

post-B.Sc. diploma course in Computer Science with a scheduled com­ puter manpower output of 2000 per year. Delhi University launched the MCA programme in 1982-3 and other four universities followed suit. In consultation with the Ministry of Education, the introduction of an eighteen-month DCA (Diploma in Computer Applications) course at the polytechnics was also planned. The programme was to be introduced at sixteen centres.32 The feasibility of introducing training courses at the Industrial Training Institutes (ITIs) in two trades, 'Console-Operatorcum Programmer-Assistant' and 'Data Preparation Assistant' were ex­ amined in consultation with the Director-General of Employment and Training (DGE&T). As the progress of computerization was slow be­ cause of the policies which prevailed prior to November 1984, the manpower problem never assumed alarming proportions before this time. However, with the liberalization of computerization policies from 1985 onwards, the manpower problem began to take on serious dimensions. Sampath headed another committee to look into the requirement of computer manpower during the period 1985-90. The Electronics Com­ mission was told about the shortage of computer manpower in 1986. The position as summarized can be seen in Table 10.6. Though the Planning Commission felt that the build-up of electronics and computer manpower through training programmes was not the DOE's responsibility, and even though education was a 'state subject', it became necessary for the DOE to act as a catalyst. Table 10.6 Computer Manpower Requirements (1985-90) Qualification

Ph.D M.Tech. B.Tech. MCA DCA DCE Vocational training

Sampath Committee Report

Revised estimates by the CCI Wing

Output from existing programmes

450 6,400 3,800 6,200 1,100 3,200

450 6,000 6,000 8,000 >40,000 3,000

NA 1,800 3,500 3,000 10,000 1,000

NA

>60,000

25,000

SOURCE: S. N. Zindal and P. K. Chaturvedi, 'Electronics Manpower Development Programmes', I PAG Journal, vol. 14,10, New Delhi, July 1987, pp. 592-3.

252

India and the Computer

The Seventh Plan allocated Rs. 10 crores to the DOE for computer manpower development. The actual expenditure was of the order of R« 13.50 crores. A paper in the

IP AG Journal summarized the proposals for computer

manpower development being undertaken by the DOE, the UGC and other agencies. The paper noted:33 The existing programmes with their natural growth rate are expected to meet less than 20 per cent of the (forecast) requirements. Any attempt to enhance the academic base to increase the output to the required levels is constrained by the lack of availability of teachers of appropriate competence. The strategy for computer manpower development would therefore have two components: (i) expanding educational base and increasing the output, and (ii) increasing the number of teachers of appropriate competence. The paper also listed of the institutions to be covered during the plan and some of the new programmes which were to be implemented.34 The total budget outlay (for the UGC, the Ministry of Education, etc.) worked out to Rs. 31 crores. The paper further noted:35 In view of the large gestation period of computer education as well as the lack of availability of teachers and limited financial resources, the estimates at this stage indicate that the output will fall short of requirements. Ways and means have to be found to plug the large gap at all levels. New approaches like computer-aided instructions, in-house training, participation of private sector etc., are being considered. The 1986 Software Policy envisaged the setting up of four Indian Institutes of Informatics Technology and promised support to private educational institutions, including the recognition of their certifi­ cates/Diplomas if they met the standards prescribed by the DOE. How­ ever, the DOE did not carry out these promises because of clashes with the Ministry of Human Resources Development which had the primary role in the field

of education. The All India Council for Technical

Education Act (15 December 1987) gave the full responsibility to the AICTE. The Study Team for the Eighth Plan made first order estimates (see Table 10.7) of the necessary additional manpower requirements and the output available from the existing institutions based on the forecast growth of the information technology sector. The Study Team felt that the shortfall at the lower levels (the DCA, DCE, and Vocational courses etc.) could be met by private institutions.

Technology and Applications Development

253

Table 10.7 Eighth Plan Manpower Requirements and Availability Level A.

Ph. D. M.Tech.

Requirement

Supply

250

240

14,500

B.Tech.

24,000

500 7,000

MCA DCA

40,000 99,000

18,000 80,000

DCE

18,500

7,000

B.

Vocational courses

6,70,000

C.

Short-term courses

16,40,000

SOURCE: Government of India, Department of Electronics, 'Estimates of Man­ power Requirement and Supply during Eighth Plan', Report of the Working Group, Eighth Five Year Plan, (1990-5), Electronics Industry, New Delhi, August 1989, p. 134. But the large shortfall in other categories would require determined government action. Otherwise the growth of the sector would be stunted. There are a large number of private computer schools imparting train­ ing. The standards are generally poor. Misleading advertisements have led to protests and questions raised in parliament. Some private organiza­ tions with political clout have taken to arm-twisting, persuading the government machinery to accord them recognition. They have managed to import PCs through their connections and have even demanded government patronage as a right. It would appear that the present scarcity of trained professionals in the computer area is a veritable boon to unscrupulous organizations which have very little merit in their teaching or training programmes. The government can be faulted for going slow, or for non-performance, but sympathy for the loud protests of private sector teaching institutions should be tempered with an awareness of their failure to setting proper standards. Perhaps they could set up their own machinery with the help of MAIT, NASSCOM and the CSI in order to police themselves. The size of the 'training industry' has been estimated at about Rs. 100 crores. A Dataquest salary survey published in April 1989 noted that the

254

India and the Computer

average salary levels of computer professionals was around Rs. 8,000 per month and the average age about 38 — lower than professionals in other industries. The remuneration offered abroad is significantly higher. A recent advertisement for posts in the UAE (for a recruitment by A.F. Ferguson & Co.) offered salaries of Rs. 14,000 with benefits and without taxes. One cannot blame the young engineers if they accept such assign­ ments in order to get ahead. Microsoft offered US $ 40,000 plus re-loca­ tion expenses and a green card to Indian Software engineers. Under these circumstances, it is difficult to prevent brain drain. Mahabala has noted that it was difficult for India to retain even 10 per cent of its IIT graduates. But some who go abroad do return, and this may help in the longer term to provide the international contacts so necessary for export marketing. For its various defence-related projects, the Defence Research and Development Organization (DRDO) requires a large number of software engineers and computer specialists. A specific programme to train 1250 graduates within a period of five years was set up in various universities. The Scientific Adviser had felt that rural talent should be tapped and hence decided that universities located away from major urban centres should be chosen for the task. A committee in 1983-4 had recommended the VAX 8350 as the best choice for the universities but for various other reasons the CDC 830 had to be used, and eight universities — Cochin, Vizag, Madurai Kamaraj, the J.K. Institute (Allahabad), Indore, Mysore, and Poona have each been provided with computer systems at a cost of Rs. 55 to 60 lakhs each. The universities are to be part of a programme aimed at meeting the special requirements of the DRDO. The Scientific Adviser also plans to subcontract some software development work to these universities as well. At least one major user of computer manpower appears to have realized the current manpower supply difficulties and set in motion a plan of self-help. The area of manpower development for the computer and allied in­ dustrial and technological fields will have to continue engage the atten­ tion of all the concerned government departments and ministries. Though it has been side-lined by the Ministry of Human Resources Development, the DOE should continue to highlight the problem and press for action. While the government may provide funding for graduate and post­ graduate education and for some vocational training programmes, private educational institutions have a major role to play in meeting the large manpower demand. The All India Council for Technical Education and the DOE have recently started an accreditation scheme — DOEACC. They must be strict in their evaluation of institutions when awarding

Technology and Applications Development

255

recognition and should not succumb to political pressures. Industry and computer users must come forward to help themselves instead of looking to the government. MAIT and NASSCOM should also take interest in this area, helping the growth of computerization through positive measures.

THE APPROPRIATE AUTOMATION PROGRAMME AND INDUSTRIAL ELECTRONICS PROMOTION In 1975, the DOE conceived the National Automation and Control Programme to be run by a National Automation and Control Agency (NACA). This was set up to create an appropriate development and production base in the fields of Control, Automation and Instrumenta­ tion .36 The Planning Commission approved the proposal as a Fifth Plan project with a total outlay of Rs. 1.8 crores over five years. The IPAG was entrusted with the responsibility for initiating this activity with a National Automation and Control Programme (NACP). The acronym did not last long, and in the Annual Report of the DOE for the year 1976-7, the project became retitled 'Appropriate Automation Promotion Programme'. The word 'appropriate' used in this context was intended to mollify those opposed to automation. Though the programme was short of money, it was not short on rhetoric. The DOE's 1976-7 Annual Report reads:37 A Systems Engineering Labora ory ... has been set up. Under the project implanted cells will be set up in suitable organizations under the joint control of the DOE and the host organization. Details of these implanted cells regarding the functional aspects, financial investments, transfer of know-how developed etc., are being finalized. During the year an expenditure of Rs. 1.5 million has been approved. The programme became part of the Industrial Electronics Promotion Programme (IEPP), and was so noted in the Annual Report of the DOE.38 The IEPP covered other projects as well, including the Microprocessor Application Engineering Programme and the Mining Electronics Equip­ ment Development Programme, both of which were started later on. The reports for the years 1978-9 and 1979-80 are voluble about the activities of the IEPP, of 'implanting' cells in various organizations and in acting as consultants for technology tie-ups — where perhaps, their expertise was less than that available in the large public enterprises which had a better grasp of the technological requirements of their projects. The funds deployed in the initial stages were most modest. The DOE

- 8

»^

—

5? 8 8 8 8 o

8 o

E—o o *-« |•

8

R

3

8 d

8 o

•2 S g

8 8 8 8 8 8 8 r4 i-5 o

8

2 8 8 2 2 8 8 r-i o o cn -»t o o

8 o

~

$ 2 S S 2 8 8 o o r4 ^ o o

8 o

s s 8 8 8 8 «-5 o o o o o

8 o

K 8 8 » S 8 o o ci d o o

8 o

I| c * |lg *** fl»

E E

2oo o-2 c .o o E

•s E—

2 ft.

r—*

1—

8

8 o rt-

o co

CO «o

co

i/"i

o

»—i

o

«/S r» CM

CM o

in t—«

8. o >>•£ 00' o O c. a. £ 0 o « c c .2? 5 c U c1.2 .2 s 2 £2 ^c 3 E "S I E co c 00 2Q- Q c « 5 o I"g •a c « co


E Ew 2 a.

3 U

—

'5. E o

g.g

00 CO CO

TO

u Q U

I

U

tii O

.O ra CO

C C a: I 8. c |5 3 a

"co o

H

u

©

o

CO

India and the Computer foreign collaboralions, so the technology development activities of the government had little direct impact on the industry in general Out of a total expenditure of Rs. 275 crores, outlays on the National Informatics Centre absorbed Rs. 139 crores. Mainframe manufacture, the C-DAC programmes, and the KBCS work, should have linkages between them established early on under a unified organization so that some specific end products are manufactured. ould one hope that the next mainframe, an artificial intelligence based supermini, or a parallel processing supercomputer could result from these combmcd efforts, obviating imports ? What befell the ECIL effort in the 70s should not be allowed to happen again in the area of mainframes and supercomputers. Manpower development, though not the responsibility of the DOE, has rightly received attention. Over Rs. 48 crores has been spent on these schemes. Whether the DOE wtndow should remain open, or whether the Ministry ofHuman Resource Development can assume full responsibility has to be decided. The educational institutions seem to prefer the DOE window because of their closer interaction and technical interest How­ ever, many of the programmes tend to overlap, though perhaps this is inevitable in a multi-pronged effort. In spite of good intentions software development has not received the necessary attention. Software Parks, though talked about, have been slow starters. A critical look at the other projects in the area of application promotion is necessary, before allocating Eighth Plan funds. The DOE has been able to attract talented engineers to initiate several of these projects, but the effective transfer of technology from R&D to the industry is in doubt. These projects should not be carried forward in the Eighth Plan, but their activities instead merged with the related public sector industrial en­ terprises or educational institutions. ERNET could with some advantage be trans erred to the NIC or the CMC Ltd., in order to give direction to the development programmes. The DOE node of the ERNET could then in a 8 ° °f 8 technological environment. has been generous in its provision of providing assistance to India for computerization. Table 10.14 gives a fair idea of the involve­ th

C

w o r k i n

o d

hc UNDP

ment o the UNDP in promoting computerization here. This assistance has been useful in providing sorely needed funds for academic work in various areas. Equipment, travel grants, fellowships, and the provision of experts to conduct various courses, has all helped in promoting a com­ puter culture in India.

Technology and Applications Development

271

Table 10.14 UNDP Assistance for Computer Technology Development (US $ million)

NCSDCT National Informatics Centre Project INTERACT Computer-Aided Design Computer-Aided Management Appropriate Automation Programme Computer networking Knowledge-based computer systems Microprocessor Application Programme Computer system for SAIL NC/CNC training (ILO) Vocational training (ILO) Total

3.00 4.40 2.75 1-74 1-28 0.84 6.00 5.25 1-53 1-06 0.70 1-08 29.63

SOURCE:

Compiled from various UNDP Reports.

NOTES 1.

Estimates Committee, 'Department of Electronics', 1973-4, Fifth Lok Sabha,66//i Report, New Delhi, April 1974, p. 143, para 5.17, and appendix 1, p. 214.

2.

Ibid., pp. 143-4, para 5.18.

3.

Public Accounts Committee, 'Computerization in Government Depart­ ments: Department of Electronics', 1975-6, Fifth Lok Sabha, 221st Report, New Delhi, April 1976, pp. 298-300, paras 7.79, 7.80, 7.88.

4.

'Technology Development Plan for Computers', IPAG Journal, vol. 3,11, New Delhi, August 1976, pp. 902-8.

5.

Ibid., p. 906.

6.

'The CAG flays IARI for poor research management', The Economic Times, 14 May 1990.

7.

Government of India, Department of Electronics, Annual Report, 1981-2, New Delhi, 1982, p. 125.

8.

Letter no: RCC/CH/26/784/87-8, dated 28 January 1988 from Dr D.R. Taneja, Director, RCC, Chandigarh.

9.

'Report of the Study Team on Computers/Office Equipment including Software', IPAG Journal, vol. 12, 6, New Delhi, March 1985, p. 301.

India and the Computer 10-

°f

11.

Summary of Activities, NCSDCT 1974-9, TIFR, March 1979.

12.

Government of India, Report of the Review Committee on Electronics (Sondhi Committee), New Delhi, September 1979, p. 72, para 12.32.

I3'

Se„CHMayTmRI "*

n

< Economic

14.

Government of India Department of Electronics, National Centre for ^oftware Technology (NCST), Annual Report, 1985-6, New Delhi, 1986,

15'

m^CnCCOUn,S C°mmiUee' 'Computerization in Government Depart­ ments^ Department of Electronics', 1975-6, Fifth LokSabha, 227s/Report New Delhi, April 1976, p. 125, para 4.22. '

16'

NewCDeThi"t1977"p'ai7oaPar,n1enl°''»»-7.

l7'

18'

ST*?'1""""'J5epar,raem °f Elec,ronics' Centre, Status Report, New Delhi, May 1983.

19.

Business Standard, 5 January 1985.

20.

Informatics

'Supercomputer 'undesirable' - Centralisation increasing: Hegde The

Economic Times, 31 July 1986.

21'

Z™r RCOf "r Co,n™li"ee Finings' and 'CRISP Comes into its own in Rural Backwaters , Computers Today, May 1989, pp. 28-33.

22.

Michael L Dertouzos and Joel Moses, The Computer Age: A Twenty Year eview, 1979, MIT Press, Cambridge, Massachusetts, USA, p. 448.

23.

It is skill that counts', The Hindu, 17 January 1988.

24.

NCERT, Report of the National Workshop on Computer Literacy Curriculum, 26-7 March 1984, New Delhi, pp. 26-30.

25.

Alison B. Bass, Span, USIS, November 1987, p. 30.

26.

27

TAG pulls up NCERT on the CLASS project', DH News Service, The

Deccan Herald, 26 September 1991.

Government of India, Electronics Commission, 'Manpower', Perspective Report on Electronics in India, New Delhi, June 1975, pp. 655-65.

Technology and Applications Development 28

273

'Report of the Panel on Computer Manpower Development', IPAG Jour­

nal, vol. 8, 2, New Delhi, November 1980, pp. 55-70, 29.

Government of India, Department of Electronics, Annual Report, 1980-1, New Delhi, 1981, p. 43.

30.

Ibid.

31.

Government of India, Department of Electronics, Annual Report, 1981-2, New Delhi, 1982, p. 37.

32.

Government of India, Department of Electronics, Annual Report, 1983-4, New Delhi, 1984, pp. 46-8.

33.

S.N. Zindal and P. K. Chaturvedi,'Electronics Manpower Development Programmes', IPAG Journal, vol. 14, 10, New Delhi, July 1987, pp. 592-3.

34.

Ibid., p. 593.

35.

Ibid.

36.

Government of India, Department of Electronics, Annual Report, 1975-6, New Delhi, 1986, p. 93.

37.

Government of India, Department of Electronics, Annual Report, 1976-7, New Delhi, 1987, p. 184.

38.

Government of India, Department of Electronics, Annual Report, 1978-9, New Delhi, 1989, p. 91.

39.

Government of India, Department of Eleectronics, Annual Report, 1984-5, New Delhi, 1985, p. 22.

40.

Government of India, Department of Electronics, Annual Report, 1980-1, New Delhi, 1981, p. 119.

41.

Government of India, Department of Electronics, Annual Report, 1987-8, New Delhi, 1988, p. 35.

CHAPTER XI

The Private Sector Takes the Lead

The formation of the Electronics Commission and the setting up of the Department of Electronics in the 1970s led entrepreneurs in India to believe that the government would encourage and support their activities. But this was not to be. Mcnon was compelled by the prevailing cir­ cumstances to support ECIL as the national champion and would not easily approve technology acquisitions or collaborations, and he relied solely on indigenous development to usher computerization into India. His pressure on IBM to conform to FERA found him an ally in ICIM, who were themselves concerned about IBM's penetration into the Indian market. ICIM adopted a posture of co-operation with the government and agreed to conform to the policies and were duly rewarded with licences and permits. But ICIM, a tradition-bound company, could not exploit the opportunities given to them. Menon also favoured the large industrial houses like Tatas, Sarabhais and DCM. He granted them limited licences without any real freedom of operation. None of these large groups could take advantage of the opportunities created by the rapid technological changes taking place elsewhere. Also the the restrictive import polices of the government impeded progress. The private sector reacted with lobby­ ing for the removal of Menon and a liberalization of the policies. There were other departments of government — like the Ministry of Industry willing to help the trade and the private sector, and political support could be found to oust Mcnon. Mcnon was replaced by Nag and the new policies gave some scope for the private sector. Though there was no liberalization of the import policies in respect of CKD/SKD kits, entrepreneurs found ways of importing kits and the government turned a

Private Sector Takes the Lead

275

blind eye. By 1979, the private sector firms could make deliveries of complete microcomputer systems using imported subsystems. The per­ formance of these computers were far superior to the systems of ECIL and they cost less. ECIL, therefore, faced difficulties in the market and rapidly lost its privileged position, the market passing into the hands of the private sector. The liberal policies introduced by Rajiv Gandhi provided the oppor­ tunity the Indian traders and entrepreneurs were waiting for. Com­ puterization got the go-ahead. Seshagiri dressed up the policy of import under the guise of industrial promotion. The new approach was timely. The technological change brought about by the IBM PC and their clones, the MS-DOS operating system, and standard software packages, could be exploited and the world computing trends brought to India. A large number of computer scientists and computer technicians found employ­ ment. The agent of change was not the software of NCST but the MS-DOS operating system and the related popular software packages, like Wordstar, Lotus, dBase and others. Rajiv Gandhi's push and Seshagiri's dynamic execution will be remembered as significant landmarks in the history of computing in India. The private sector companies — HCL, Wipro, Sterling and others soon established connections with the international trade and industry, developed vigorous marketing abilities, and were able to exploit the policy loopholes of the protected Indian market. Shivnadar left DCM in 1975 and took five others with him to start Hindustan Computers Ltd. Following the US tradition of cntreprencurship this cohesive team built up HCL. With an intimate understanding of the markets and with the knowledge that the microcomputer can reduce computing costs, they pushed ahead and established themselves as the market leader. Other entrepreneurs like A.H. Prcmji, Chairman of Wipro, and C. Sivasankaran of Sterling followed. The success of the new companies promoted by young entrepreneurs is a pointer to what enterprise can achieve, suggest­ ing that if the government allows the free play of market forces in a regulation-free environment, the market will boom, creating rewarding opportunities for the young and the talented. Hewlett-Packard (HP), has now made arrangement with HCL for marketing their products in India and for providing service support and has taken a significant equity interest in HCL—in fact a new joint company has been set up. This joint company HP-HCL could become the market leader in India. DEC, in association with Hinditron, has estab­ lished Digital Equipment India Ltd. and is gradually improving its market

—

u

£ £ £

' A O ' O O N O ^ V i C Q o O l/l D O m m . ^ n > c « n r i o o ^ « £ 2g £ g £ g £ P 22 3 2 "• !C 2 2 t ^ ^ rsl£^Z5500v°',1'/^'^

« N O N N © t co r- r-

1

oo cj © d oo vo' d r4 pJ f t W N N I N NH (N

o ec o v oo "i r- q o o-> — ® ^ t ^ N N O h O O v o o o d o o « o oo —' © >o d d r d K r i fN •— — — — (N

©P2

.2 £

.E E E 2 S.-5 a ™

£ o

o ~ £

5 o ~o oo roo ro

f rf*0 >— — —

>© >© Tt

£ 5

s

S . J

I s

« 8. £ s. s

O rs| — — T

s

T — o O n —

v©

o

d

d r~

d n

rg rs oo \o oo >©

r-

©• >©

•—

d

oo 00 •—•

d

v© wo ao «*>

T

O •© oo SO •© — d — 00. d ro —'

©

r-. d"

oo = O *

Q

.2 c

O S 7 = H u CQ r J i . ^ 2 j w S 2 l « j S s | g J s ^ ^ o m U c , u J J 2 > ^ - u n i i ™ ' J >u 5 & O.E < f s j ; I > U u J a . ( - _ Q v 5 f - Z ( - Q N u S l JO £

Q

Private Sector Takes the Lead

277

penetration. Olivetti of Italy in collaboration with Modis have introduced various versions of their PCs into the Indian market. Texas Instruments has set up a software export operation at Bangalore. The private sector base in computers has been considerably augmented by the arrival of these MNCs on the Indian scene. Dataquest brought out an informative set of statistics in respect to the computer industry in its July 1991 issue and the following tables com­ piled from this represent the current status of the industry.1 Table 11.1 lists the top twenty industry leaders. Table 11.2 sum­ marizes the Information Technology scene in India. The total value of goods and services appears to exceed Rs. 2,213.89 crores. At interna­ tional prices, the value may be about Rs. 900 crores and the import Table 11.2 Information Technology Activity in India - A Rough Outline (1991) (Rs. crores)

1.

Microcomputer PC\XT\AT\386\486

495.30

2.

Minicomputer including supermicros

153.78

3. 4.

Workstations/CAD Superminis and mainframes

116.16

5.

Other items Total hardware sales

94.82 45.01 905.07

6. 7.

Consumables Maintenance

135.42

8.

Special projects

9. 11.

Peripherals Software

12. 13.

SEEPZ Exports Hardware exports

178.88 68.60 187.33 151.71 65.07

14. 15. 16.

Software exports Hardware imports Training

235.45 25.00 86.36

Total IT Business SOURCE: p. 18.

175.00

2,213.89

'Indian IT — Value of Goods and Services', Dataquest, July 1991,

India and the Computer content at least Rs. 750 crores. The value addition in hardware rem*' poor because of the lack of depth in manufacture.

S

com 'e™ J"h 2 tUr"OVer 0f OVer Rs- 218-35 CTOrespublic sector company CMC leads in systems design, software and seiwices. The two Tata companies lead in software. The share of the public sector as a supplier of goods and services to the information and technology sector may no. exceed 25 private sector has forged ahead thanks to the change in government polices, but the government and the public sector remain the dominant —S

'hC 8°°dS a"d SerViCeS being pr0Vided hythe Priva'e sector

and * • , P 'Wenty flrms',he others have ve[y small turnovers nd their survival as manufacturers is, therefore, doubtful - they may to da e°exWceed 5°o!Th °'T enginee"^ The ^"ces/regis,rations e"S'ng and licensinS P°licies moan very nidi tie uds1 d'S ~l °" en,erprise' a '"hnological orientation, techfnA' h a marketing skills will lead to results. The history of the SEEKS!'iCCnSin& ^ °,h" — ~• —

IBM®!?. bUf"eSS and imp°rtS have also boomed in lhis periodIBM s sales alone amounted to over Rs. 40 crores in 1988- and the total

- — software imports 89 90 and to Rs. 25 crores in 1990-1.^However,the crores in 198

9

d

d to

ZTZ:::; ""R1-•»»

iustcd them g*"3 c/i

q NO CM r-^ iri 00

Q-J

00 00

r\

S s i

c S ^.2 G oo — 2£ g

rn ON ON «o °i. n ©N Tf ©_ ^ c> CM CM* «n* M-* r-* CM* o CL CL

•j

j> •O

'>

C/3

ao c i7) CA U 2 O 22 3 Q. 2 E « o Q U

73

-

>

c "C

1 I

2 ^ 22 3 O. E o U

£ w C « •* -2 D

» z> a

x: Z O s

CM 00 o Nf 00*

Hi

o

c M n i 22 2 o la CL CL 'C « a o a.

"D 5

a>

> CO c/5 i

M" r—* r-< *—